Organisation for Economic Co-operation and Development GOV/SIGMA(2021)3 Unclassified English - Or. English 14 October 2021 PUBLIC GOVERNANCE DIRECTORATE SIGMA - Support for Improvement in Governance and Management Organisation of public administration: agency governance, autonomy and accountability SIGMA Paper No. 63 Good governance of public agencies requires the application of a set of regulatory and managerial tools to find the right balance between autonomy of agencies and adequate oversight from portfolio ministries and other actors. This paper provides insights from EU and OECD good practices, with a detailed analysis of EU acquis requirements for national regulatory agencies. New empirical evidence shows that public administrations in the Western Balkans and European Neighbourhood area lack clear policies and regulations for agency governance and misinterpret the EU acquis. This leads to a proliferation of agencies, duplication of functions and waste of public resources, a lack of accountability to portfolio ministries and generally a governance vacuum. Implementation of government policy is blocked and democratic accountability generally undermined. Finally, recommendations for better organisation of public administration are provided, based on the empirical analysis and lessons learned from SIGMA's engagement in such reforms. For further information please contact [email protected]JT03482987 OFDE This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area.
75
Embed
SIGMA - Support for Improvement in Governance and ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Organisation for Economic Co-operation and Development
GOV/SIGMA(2021)3
Unclassified English - Or. English
14 October 2021
PUBLIC GOVERNANCE DIRECTORATE
SIGMA - Support for Improvement in Governance and Management
Organisation of public administration: agency governance, autonomy and accountability
SIGMA Paper No. 63
Good governance of public agencies requires the application of a set of regulatory and managerial tools to find the right balance between autonomy of agencies and adequate oversight from portfolio ministries and other actors. This paper provides insights from EU and OECD good practices, with a detailed analysis of EU acquis requirements for national regulatory agencies. New empirical evidence shows that public administrations in the Western Balkans and European Neighbourhood area lack clear policies and regulations for agency governance and misinterpret the EU acquis. This leads to a proliferation of agencies, duplication of functions and waste of public resources, a lack of accountability to portfolio ministries and generally a governance vacuum. Implementation of government policy is blocked and democratic accountability generally undermined. Finally, recommendations for better organisation of public administration are provided, based on the empirical analysis and lessons learned from SIGMA's engagement in such reforms.
This document has been produced with the financial assistance of the European Union (EU). It should not be reported as representing the official views of the EU, the OECD or its member countries, or of partners participating in the SIGMA Programme. The opinions expressed and arguments employed are those of the authors.
This document, as well as any data and any map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area.
I. From agencification to consolidation – trends in organisation of public administration 9
“Agency fever”: international trends and their impact on transition economies 9
Back to consolidation? The de-agencification trend 12
Gatekeeping functions and regular reviews 15
Summary - Agencification is not a panacea 17
II. Supranational standards – towards the EU model of national agencies 18
National regulatory authorities – the flagship “EU-made” national agency 18
Expanding the standards of autonomy to non-regulatory bodies 30
Summary – EU standards for national agencies 32
III. International experience and tools for agency governance 33
Ministerial management and oversight tools to balance agency autonomy and control 33
Developing a culture of ministry-agency collaboration and clarifying roles 41
Summary - Good practices and tools for agency governance 43
IV. Agencies in the Western Balkans and the European Neighbourhood– how autonomous? How accountable? How to reform? 45
Introduction 45
Agencification in the Western Balkans and the European Neighbourhood – an overview 45
How autonomous, how accountable? 52
How to reform? Lessons learned from SIGMA support 63
Summary 67
References 69
The SIGMA Programme 75
4 GOV/SIGMA(2021)3
Unclassified
FIGURES
Figure 1. Dimensions of agency autonomy 11 Figure 2. Observed strategies for clustering public agencies 15 Figure 3. Expansion of EU law standards on organisation and autonomy of national regulatory authorities 20 Figure 4. Layers of agency accountability 34 Figure 5. Building blocks of the ministerial governance framework for agencies. The case of oversight
agreements in Ireland 37 Figure 6. Classic mechanisms of control of agencies by portfolio ministries 41 Figure 7. Summary of good practices and tools for agency governance 44 Figure 8. Number of public agencies in the Western Balkans and the European Neighbourhood 46 Figure 9. Number of agencies subordinated to the parliament in the Western Balkans and the European
Neighbourhood (constitutional bodies excluded) 50 Figure 10. Results-oriented governance of agencies in the Western Balkans and the European
Neighbourhood 52 Figure 11. Key characteristics of agency autonomy 54 Figure 12. Bodies approving annual plans and setting targets for agencies 57 Figure 13 Use of performance auditing by portfolio ministries 58 Figure 14. Ministerial feedback mechanisms on agency performance 59 Figure 15 - Types of feedback provided by portfolio ministries to agencies 60 Figure 16. Use of binding instructions by portfolio ministries 61 Figure 17. Allocation of responsibility for oversight of agencies within portfolio ministries 62 Figure 18. Agency involvement in policy making 62 Figure 19. Imbalanced autonomy and accountability of agencies in the Western Balkans and the European
Neighbourhood 63
TABLES
Table 1. EU acquis standards of autonomy of National Regulatory Authorities 21 Table 2. Multi-sector regulators in EU member countries 24 Table 3. European Commission guidelines on interpretation of independence standards for electricity and gas
market national regulators 25 Table 4. Clarifying the NRA’s autonomy. What is and what is not required by EU law 25 Table 5. In search of equilibrium: calibrating ministerial management of agencies 33 Table 6. Typologies of public agencies in selected European countries 35 Table 7. Selected European countries with framework laws – examples of ministerial management
instruments 36 Table 8 Existence of policies and regulations to manage central government agencies 48 Table 9 Clarity and comprehensiveness of official typology of central government agencies 49 Table 10 National regulatory authorities reporting to the Parliament in the Western Balkans and the European
Neighbourhood 51 Table 11 Agency autonomy – differences in views across agencies and portfolio ministries 53 Table 12 Ministry-agency interaction – differences in views across agencies and portfolio ministries 56 Table 13 Summary of key deficits in agency governance in the Western Balkans and the European
Neighbourhood 68 Table 14 Dos and don'ts of agencification - key recommendations 68
GOV/SIGMA(2021)3 5
Unclassified
The governance of public agencies, in all their forms, matters greatly for the overall performance of the
public administration. The rules established for their creation, operation and possible termination by
political actors, and enforced by centre-of-government institutions and ministries, determine the
architecture and overall organisational landscape of the public administration. These rules should also
specify the autonomy needed for each type of public agency to function well, and the accountability
mechanisms needed to ensure that the agency is held to account for its performance and that it uses public
funds effectively and efficiently for their intended purpose. Form follows function and some types of
agencies, especially regulators, need an arms-length relationship with ministries to perform their duties.
The actual levels of autonomy and accountability – and the overall institutional performance – depend on
the application of a set of management mechanisms by portfolio ministries or other actors responsible for
overseeing the work of agencies.
Public agencies exists in all countries, in rich variety, but historically the pendulum has swung from
consolidated organisational structures with larger ministries towards more specialised public agencies with
responsibility primarily for public service delivery. However, many European Union (EU) and OECD
countries have reversed agencification trends in recent years, bringing back core public functions to the
ministerial hierarchy, as the negative effects of the initial wave of “agency fever” became clear. They have
also developed better regulatory and managerial tools to ensure that the increased autonomy of agencies
was balanced with the necessary oversight. Reaching this equilibrium is a continual challenge.
This paper shows that the pendulum has never swung back for a large group of transition economies:
those of the Western Balkans and the European Neighbourhood. In fact, in many countries and policy
areas it has accelerated and created imbalances between autonomy and accountability mechanisms.
These transition economies embraced agencification to unprecedented levels, but without applying the
same regulatory and managerial tools that the OECD and EU countries realised were needed to prevent
functional irregularities. Since the 1990s, these countries have experienced massive agencification, with a
multitude of agencies created overnight to replace monolithic, heavily-centralised administrations inherited
from socialist regimes. Governments did not have two or three centuries to incrementally craft and polish
government-agency relations, as was the case, for example, in Scandinavian countries. International
partners often promoted agencification in transition countries, without considering that this organisational
model was not part of their administrative DNA.
This paper presents empirical evidence to illustrate that agencies in general have excessive levels of
autonomy and operate in a governance vacuum in most of these transition economies, with little
management or oversight from portfolio ministries. A specific type of agency, referred to as national
regulatory agencies in the language of the EU acquis, has such extreme levels of autonomy that they often
operate as a fourth branch of government, not accountable to anyone in practice.
Unfortunately, this is not a unique situation. Studies have previously documented the same pattern of
unaccountable public agencies and the associated policy outcomes in the countries in Central Eastern
Europe that would later become EU Member States. However, this empirical evidence was at the level of
individual case studies. New data from SIGMA Monitoring Reports and a unique survey conducted of 236
senior managers of public agencies and their counterparts in portfolio ministries provide the first systematic
documentation of this regional trend that has far-reaching implications for public governance. In short, the
Executive summary
6 GOV/SIGMA(2021)3
Unclassified
proliferation of agencies combined with a governance vacuum has led to clear duplication of functions and
waste of public resources, blockages for implementation of government policy and thereby also a general
undermining of democratic accountability. In the worst cases, agencification is used for political spoils and
the capture of institutions, retrenching patronage patterns. This paper identifies the following adverse
effects of uncontrolled and unmanaged agencification in the Western Balkans and the European
Neighbourhood:
Lack of accountability to ministries, where classic executive agencies are effectively isolated from
the government administration and formally subordinated to the parliament. The most extreme
case is Kosovo*, with over 30 agencies reporting only to the Parliament. It not only hinders the
government’s capacity to oversee policy implementation, it also makes the supervision of agencies
largely fictitious, as the parliament has no capacity to perform these functions properly.
Many administrations do not have a clear policy for how central government should be organised,
and what role agencies play. Ministries are often too weak, both in terms of administrative capacity
and political leverage, to initiate coherent organisational reforms and ensure a consistent approach
to governing agencies across the government. Different approaches to organisation can therefore
be observed in different parts of the administration, as legacies persist from previous times.
Fragmented regulatory framework for management of agencies. While most administrations have
framework laws on public administration in place, these acts are usually bypassed by special laws
regulating the individual agencies and creating different arrangements for them. Framework laws
fail to establish comprehensive governance frameworks for agencies promoting results-oriented
management.
Misinterpretation of the EU requirements on autonomy of specific types of agencies, such as
regulatory authorities. It is often understood (incorrectly) that it is a requirement to make these
agencies accountable solely to the legislature.
Lack of safeguards against proliferation of agencies. The ‘new functions – new agency’ logic is
widespread and new agencies are created without thorough analysis of alternative options for
delivery of specific functions. Subsequently, their performance is not subject to regular review or
‘existential tests’.
Reforms are not always successful but are nevertheless necessary. Gatekeeping functions and careful
reflection when creating new structures is essential, as merging or abolishing an agency has proven almost
impossible once it is created, at times because it is a part of larger political agreements. Frequent changes
of governments make long-term reform initiatives difficult to manage at a technical level, and the national
administration often does not have the capability to successfully drive whole-of-government reform
initiatives. The EU legislation in this domain is often misunderstood, sometimes deliberately misread. This
paper therefore clarifies the requirements of EU acquis for each type of agency. Sequencing of reforms,
technical support, broad coalition building, high-level political engagement and a good deal of
perseverance are key to the success of reforms.
This designation is without prejudice to positions on status and is in line with United Nations Security Council
Resolution 1244/99 and the Advisory Opinion of the International Court of Justice on Kosovo’s declaration of
independence.
GOV/SIGMA(2021)3 7
Unclassified
The organisational setup of public administration matters for its performance. The Principles of Public
Administration recognise the importance of this issue by stating that the overall organisation of central
government should be rational, based on adequate policies and regulations and provide for appropriate
internal, political, judicial, social and independent accountability (OECD, 2017[1]). Recognising the
fundamental importance of effective institutions, the European Commission asked SIGMA to develop this
paper to provide advice to governments in the Western Balkans and the European Neighbourhood. The
paper summarises the requirements stated in EU legislation regarding organisational structures,
specifically relating to levels and forms of autonomy, for key policy areas. It complements these standards
with OECD standards and guidance from several legal instruments and policy papers.1 Finally, the paper
reflects on the experience of reforms in the Western Balkans and the European Neighbourhood and
provides recommendations.
The organisational landscape of public administration has been transformed in recent decades, as tasks
traditionally performed by ministries were delegated to public agencies, state-owned enterprises or the
private sector. This general trend towards less centralised management was expected to bring
considerable efficiency gains. However, it also created challenges for the effective governance of
administrative structures that are more heterogeneous and complex than ever, as shown in Chapter 1.
International standards have evolved, with the EU model of national regulatory authorities as the most
prominent example of supranational standards, as discussed in Chapter 2. OECD and EU countries have
learned many lessons for agency governance from the early waves of agencification. Chapter 3 presents
this international experience and the tools that transition economies can also benefit from to strike a good
balance between autonomy and accountability.
In Chapter 4, the paper combines the insights from SIGMA’s past and ongoing support and evaluations in
the Western Balkans and the European Neighbourhood with new evidence on how the practice of agency
governance in these economies corresponds with international trends and standards established by EU
law and OECD instruments. The focus is on regulatory authorities and executive agencies. EU legislation
sets standards for both types of agencies, but in particular for regulators. However, these standards are
often misunderstood or misapplied in the Western Balkans and the European Neighbourhood. The analysis
draws on past SIGMA evaluation reports, review of the legislative framework, official administrative data
as well as a unique survey of more than 273 senior managers of public agencies and senior civil servants
in ministries responsible for oversight of these agencies in ten Western Balkan and European
Neighbourhood governments.
Chapter 4 ends with recommendations on managing the organisational setup of central public
administration. The main objective is to help governments effectively oversee and manage their agencies,
while respecting their functional independence when required. The recommendations are primarily
addressed to the governments of the Western Balkans and the European Neighbourhood, but may also
be useful for international partners supporting reforms in these regions, in particular relevant directorates
1 The OECD Policy Framework on Sound Public Governance (OECD, 2020[83]) weaves together the relevant OECD
legal instruments and tools on public governance. https://www.oecd.org/governance/policy-framework-on-sound-
a wide array of agencies in all or a vast majority of policy areas.
Sectoral consolidation – reorganisation of the agency landscape within a single policy area or
within a specific type of government function.
Comprehensive consolidations were only implemented in some countries, for example in the UK and
Ireland (Box 1), whereas sector-level consolidation occurred in wide array of countries. While differing in
scope, both types of consolidation followed a similar pattern – reducing the number of agencies through:
(1) merging agencies; (2) transferring agencies’ functions back to the ministries; (3) withdrawing from
performing specific functions resulting in abolishment of agencies; or (4) delegating these functions to other
bodies, such as local governments, the private sector or civil society organisations.
GOV/SIGMA(2021)3 13
Unclassified
Box 1. Comprehensive consolidations of agencies in the UK and Ireland
The UK’s Public Bodies Reform Programme (2010-2015) and Ireland’s Agency Rationalisation
Programme (2011-2013) followed a similar pattern, combining massive restructuring of agencies with
setting mechanisms for regular review of the number of agencies. The UK’s programme began with a
review of over 900 government organisations that led to identification of over 200 bodies that no longer
needed to exist as standalone organisations and over 170 bodies that could be merged with others.
These targets were largely accomplished. Within five years, the total number of public bodies was
reduced by over 290, thanks to the abolishment of 190 institutions and the merging of 165 others into
fewer than 70. The reported cumulative reduction in administrative savings reached GBP 3 billion. The
overarching logic behind the restructuring was to select the most effective and efficient option for
delivery of specific government functions. It is interesting that allocating functions to non-departmental
public bodies (primary type of government agency, operating in an arm’s length capacity with the parent
ministry) was indicated as “a last resort, when consideration of all other delivery mechanisms have been
exhausted” (Cabinet Office, 2012, p. 4[24]). The catalogue of alternative delivery options included listed
devolution to local government, transferring to private or voluntary sector, bringing in-house (to the
parent department) or creating an executive agency (part of the government department). In addition
to this one-off consolidation exercise, the mechanism of post-consolidation reviews (Triennial Reviews)
was introduced. All non-departmental public bodies were required to go through a review process over
a three-year cycle. The key principle of the reviews was their challenging character, i.e. they require an
'existential test’ for each body in order to analyse whether there is a robust justification for retaining
them as standalone organisations. The model of Triennial Reviews was recently replaced by tailored
reviews that rely on slightly simplified methodology. However, the general concept of regular and
challenging reviews was retained. Finally, the oversight of the existing agencies was tightened by
introduction of new requirements on transparency of salaries and contracts, as well as new forms of ex
ante external controls for specific transactions.
The Irish Agency Rationalisation Programme reduced the overall number of Government bodies by
almost 200. It also introduced a principle stating that the creation of a separate agency is only justified
when the performance of particular functions requires specialist skills and autonomy. Similarly to the
UK, one-off consolidation was complemented with the introduction of Periodic Critical Reviews (PCRs)
for all state agencies (see the box below).
Source: (Cabinet Office, 2014[25]) (Department of Public Expenditure and Reform, 2014[26]) (Department of Public Expenditure
and Reform, 2016[27]) (Dommett, MacCarthaigh and Hardiman, 2016[28]) (MacCarthaigh and Hardiman, 2017[29])
Comprehensive consolidation initiatives led to a significant reduction in the number of agencies, though in
the Irish case its impact was diminished by the creation of several new institutions in parallel. The public
agency was no longer the default option for delivery of all government functions, but only one of the options.
Sunset clauses and regular reviews of agencies were introduced to ensure that agencies continued to
serve a purpose years after their establishment. The rationalisation programmes implemented in Ireland
and the UK also challenged another constitutive element of the agencification agenda – the need for strict,
organisational separation of policy formulation and policy implementation. The benefits of increased
democratic accountability through shifting more functions under the direct leadership of elected politicians
(ministers or local government authorities), were emphasised. Some whole-of-government consolidation
initiatives were also undertaken in the Western Balkans and the European Neighbourhood, but the results
are rather mixed (see experiences from Kosovo and Albania in Chapter 4).
14 GOV/SIGMA(2021)3
Unclassified
Sectoral consolidations are diverse and frequent. Four main types of sectoral consolidation can be
identified from recent European practice (Sześciło, 2020[30]).
Amalgamation of all service delivery agencies operating within one policy area, e.g. merging
customs and tax administration (Denmark, Poland, UK), merging social services and labour
services (Norway).
Amalgamation of regulators, e.g. merger of financial regulators (Finland, Poland – merger of
financial supervisory authority and insurance supervisory authority); merger of electronic
communications and postal service regulators (Croatia, Poland, Slovakia, UK); merger of
competition authority with post and telecommunications market regulator (Netherlands). One of
the most prominent examples of such reforms was establishment of the Spanish ‘super-regulator’
- the National Authority for Markets and Competition (CNMC), set up in 2013. It replaced the
National Competition Commission (competition protection authority) and separate sectoral
regulators in electronic communications, the audio-visual sector, energy, railways and airports,
and postal services.
Integration of inspectorates (regulatory enforcement agencies). This may include integration of
inspectorates operating within a specific risk area, following the ‘one risk – one inspectorate’
approach promoted by OECD (OECD, 2014[31]) e.g. by merging all bodies handling food security
(Denmark, Germany, Ireland, Netherlands). The most radical approach was first applied in Croatia.
In 1999, the centralised State Inspectorate took over tasks of 12 specialised inspectorates (labour
and worker safety and protection, trade and market surveillance, power, mining and equipment)
operating under four ministries. It is interesting to note that this approach had been already
implemented in the Western Balkans. In 2012, Montenegro created the Administration for
Inspection Affairs, integrating under one roof specialised inspections operating in the following
areas: mining, geology, electronic communications, postal services, labour, tourism, construction,
environment, public procurement, games of chance.
Consolidation of provision of back office services for government administration under a single
roof of shared support centres (Estonia, Finland, Ireland, Netherlands, Portugal, Sweden, UK).
This includes, in particular, integration of provision of such services as property and fleet
management, financial management (accounting, payroll administration), procurement or IT
infrastructure.
Integration of territorial branches of the government administration. In Lithuania, the proposals for
rationalisation of government administration have been formulated by the so-called “Sunset
Commission” since 1999, but considerable results were only delivered during the global financial
crisis. Abolishment or reorganisation of territorial branches of government administration was the
major factor leading to reduction of the total number of government bodies by nearly 10%, between
2008 and 2010 (Nakrošis and Martinaitis, 2011[32]). In Hungary, the integration of territorial
administration under the roof of county government offices brought an even sharper reduction of
the number of administrative bodies, but surprisingly, led to an increase in employment (Gellén,
2012[33]).
Sectoral consolidations typically rely on grouping similar functions into clusters in a single organisation.
This approach has been popular for governments seeking efficiency gains and synergies through
organisational reforms. For example, in the Dutch Compact Central Government Programme (Government
of the Netherlands, 2011[34]) the cluster approach was presented as a strategy for streamlining the
Government’s operations, creating synergies and savings by co-ordination or consolidation of delivery of
similar functions. The Government identified several distinct groups, where organisational integration could
be introduced, e.g. providing income support to citizens, collecting contributions, managing publicly owned
property, inspections and market supervision (regulatory functions). For example, the identification of this
last group resulted in the above-mentioned merger of the competition protection authority with the regulator
GOV/SIGMA(2021)3 15
Unclassified
of the postal and telecommunications market. With regard to the inspection group, the idea of a single
state inspectorate was considered, although eventually it resulted in a significant reduction of the number
of agencies. An important element of the Dutch approach is the designation of the ministry responsible for
each group.
The cluster approach could serve as a general principle shaping the agency landscape. The catalogue of
groups identified in the Dutch case serves as a good point of departure, but as the examples discussed
above show, there are other potential paths towards the integration of agencies. Where the merging of
relevant agencies is not feasible, better co-ordination and common operating standards should be ensured.
The figure below presents a map of clusters that could guide the process of consolidation or greater
integration of agencies. These clusters may overlap to some extent (e.g. consolidation of revenue
administration might be combined with the merging of inspectorates operating in this area), but they could
also serve as alternative options.
Figure 2. Observed strategies for clustering public agencies
Gatekeeping functions and regular reviews
Some OECD and EU countries have developed gatekeeping functions to avoid the proliferation of new
agencies and restrictions to avoid ad hoc modifications to the carefully-designed, general accountability
and governance frameworks. A balanced policy response to these challenges consists of two elements:
Procedures for creation of new agencies to carefully consider the need for such institutions and
possible alternative organisational setups.
Mechanisms for regular reviews and adjustments of the existing agencies.
Mechanisms like these have been introduced in the past decade by countries that introduced large-scale
reorganisations of their public administrations, such as the UK and Ireland. Creation of new agencies
should not be perceived as a default option to deliver government functions. As the UK’s Cabinet Office
(2016) encapsulated: “New public bodies should only be created if there is a clear and pressing
16 GOV/SIGMA(2021)3
Unclassified
requirement, a clear need for the state to provide the function or service through a public body, and no
viable alternative - effectively establishing new public bodies as a last resort. This is to prevent any
unnecessary increase in the number of public bodies”. Any proposal for a new agency to be established
should be subject to ex ante review (a tailored regulatory impact assessment), guided by the following
questions:
Is the function or service to be provided needed? Is it required by legal or policy obligations?
Are similar functions performed by any existing bodies? If yes, what is the advantage of creating
a new body instead of allocating functions to an existing one?
Can the relevant function be performed more effectively and efficiently by bodies other than a
public agency?
The catalogue of alternative delivery options includes local governments, the private sector and the
voluntary sector, but also corporatisation of public bodies (primarily agencies), i.e. transforming them into
private law enterprises with the government as sole (or major) shareholder, or delegation of some
government functions to the judicial bodies. For instance, running public registers (e.g. company registers,
land registers) might be entrusted to courts in order to enhance the credibility of these registers.
In principle, if a new agency is needed, the default option is to have one operating within the ministry or
under its close supervision, while creation of a more autonomous type of agency requires special
justification. For example, creation of the Autonomous Administrative Authorities (ZBOs) in the Netherlands
is limited by law to the following cases:
There is a need for an independent opinion based on specific expertise.
Strict application of the rules is required in a large number of individual cases.
Participation of civil society organisations is deemed particularly appropriate in the light of the
nature of the administrative task in question (Section 3 of the 2006 Autonomous Administrative
Authorities Act).
Regular reviews of the existing agencies often consist of a two-step procedure: (1) challenging the very
need for continuation of the agency’s activities in the current form; (2) investigating the specific measures
improving effectiveness and efficiency of the agency’s activities. Reviews may be conducted by the
portfolio ministry or by the body responsible for public administration matters, e.g. the respective ministry
of public administration or the prime minister’s office. The box below presents the review scheme adopted
by the Irish Government.
Box 2. Compulsory reviews of public agencies
In Ireland, all public agencies (called Non-Commercial State Bodies) are subject to Periodic Critical
Review (PCR) at least every five years. The main objective of the PCR is to verify the need for the
existence of each agency. Subsequently, it aims to improve accountability, efficiency and effectiveness,
as well as investigate cases for rationalisation and consolidation. Performance is evaluated against the
predefined objectives and targets. Finally, the governance structure of the agency and ministerial
oversight are subject to review.
The following overarching principles should be considered throughout the review process: a)
proportionality – it should not be excessively bureaucratic or burdensome; b) timeliness – it should be
completed quickly in order not to disrupt the work of the agency; c) challenging character – it should be
robust and rigorous, ensuring that all delivery options for specific government functions are analysed;
GOV/SIGMA(2021)3 17
Unclassified
d) openness and inclusiveness – the procedure should provide all stakeholders with opportunities to
contribute; e) transparency – reviews should be made publicly available.
The responsibility for the review process is partially decentralised to the level of the portfolio ministry
(department) that is required to establish a working group to conduct the review. However, in addition
to the representative of the ministry and the relevant agencies, this group consists of members from the
Department of Public Expenditure and Reform (ministry responsible for public administration). It is
concluded with a report submitted to the portfolio minister.
Source: (Department of Public Expenditure and Reform, 2016[27])
Summary - Agencification is not a panacea
Agencification has significantly changed the organisational architecture of public administration across the
world. Public agencies have become the major vehicle for implementing policies and delivering services
by the government. Nevertheless, the initial enthusiasm accompanying the agencification waves of the
1980s and 1990s has largely faded. Strong empirical evidence has not been provided that agencification
leads to enhanced efficiency, effectiveness and quality in policy making and implementation. There are at
least three major lessons that can be drawn from the agency fever of recent decades.
First, there is no ‘one size fits all’ model for organising the public administration. In particular, there is no
formula for the optimal balance between consolidation and specialisation in administrative structures. As
stated by Pollitt: “(…) distrust those who argue that X must be implemented because ‘everyone else’ is
doing X. Actually, there are no universal solutions because there are no universal problems” (Pollitt,
2012[35]). Second, an agency is not, by default, more effective or efficient in delivering specific
implementation functions than a ministerial unit, a state-owned company, or a private sector or local
government entity, as was sometimes believed in the early days of the agencification era. Thus, an agency
should not be perceived as a default option for government functions. It is increasingly understood that the
creation of an agency should be preceded with comprehensive ex ante analysis of existing organisational
arrangements for delivery of public functions. Third, irrespective of the organisational form selected for
delivering specific functions (ministerial unit, agency or state-owned company) similar governance
challenges persist. The key challenge is maximising the benefits of delegation and autonomy, while
minimising the risk of irregularities or errors and ensuring the effective accountability of all bodies
performing government functions. This requires a consistent governance framework for all bodies
performing government functions, striking a balance between autonomy and control, freedom and
accountability.
18 GOV/SIGMA(2021)3
Unclassified
National regulatory authorities – the flagship “EU-made” national agency
Traditionally, national administrative organisation was not subject to regulation of international and
supranational4 laws, as a cornerstone of national administrative sovereignty (Egeberg and Trondal,
2018[36]). However, with the progress of European integration, EU law began to provide more and more
detailed guidelines regarding the organisational design and institutional locus of the national administrative
bodies responsible for implementation of the EU legislation. This was part of a broader structural shift in EU
policy promoting a “regulatory state” (Majone 1997), where newly-liberalised markets required new models
of regulation by a special type of administrative body with an arm’s-length relationship with ministries,
broadly labelled “regulatory agencies” (Majone, 1997[37]) For the Western Balkans and the European
Neighbourhood, alignment with these rules is an important component of the process of approximation of
their legal systems to the EU acquis. Therefore, the design of the national administrative structures cannot
ignore the requirements established at the EU level.
However, EU standards have been misunderstood and misinterpreted when applied in the Western
Balkans and the European Neighbourhood. In this Chapter, the aim is to clarify the requirements of EU
standards on the organisational setup of agencies. These considerations will concentrate primarily on the
national regulatory authorities (NRAs), as they are subject to particularly extensive regulation in the EU
acquis. This analysis will be complemented with insights from OECD instruments and guidance on
governance of regulators in general.
The principle of functional autonomy for NRAs (and some other bodies) is not a revolutionary idea. It can
be traced back to when Woodrow Wilson formulated the rule that elected politicians could influence
administrative decision making by shaping the legislative framework, but not through direct and often
informal interference in individual proceedings or decisions. It is also a pillar of classic Weberian
bureaucracy, where civil servants are bound by law, not the will of political leaders. The added value of the
EU’s focus on functional autonomy, compared to these classic concepts, is the explicit exclusion of the
power of ministries to review the NRAs’ acts and the complementing of this functional autonomy with
“autonomy bonuses” in other dimensions – legal, structural, financial and managerial.
The regulation refers to the technology of governing social systems by rules (Baldwin, Cave and Lodge,
2010[38]). In this sense, most of the public authorities could be defined as regulators, either setting or
implementing regulations. However, the label of regulatory authorities is usually attached to a narrower
group of bodies tackling market failures in selected markets, using such powers as: licensing market actors
and distributing special rights (e.g. radio frequencies); monitoring compliance of market operators with a
4 The notion of ‘supranational standards’ refers to standards established by an organisation whose bodies are provided
with the authority to set laws that are binding for its member countries, based on the powers transferred to this
organisation by the members. The European Union is the leading exemplar of this type.
II. Supranational standards – towards
the EU model of national agencies
GOV/SIGMA(2021)3 19
Unclassified
regulatory framework, including imposition of sanctions for non-compliance; counteracting distortion or
restriction of competition and abuse of dominant position by operators on relevant markets; undertaking
actions improving accessibility of services on regulated markets to various groups of users; fixing or
approving tariffs (e.g. for transmission or distribution of energy); considering complaints of the market
operators against discriminatory actions hampering access to the market or fair competition; issuing
binding instructions to market actors about specific measures to be undertaken (Sześciło, 2021[39]). As
such, regulatory authorities are sometimes distinguished from bodies of narrower, “policing” functions, i.e.
regulatory enforcement agencies performing classic compliance control functions in the form of market
inspections (OECD, 2014[31]). This includes, for example, labour, environmental or food safety
inspectorates. In practice, the demarcation line between regulatory authorities and regulatory enforcement
agencies is often blurred, as regulatory authorities often also perform inspection functions.
National regulatory agencies are regulators characterised by a single additional feature: their functions,
powers and organisational setup are shaped to a large extent by EU law. The NRAs became part of the
EU acquis setting general standards for autonomy, as this was deemed crucial to the success of market
liberalisation reforms promoted by the EU. It is believed that autonomous regulators are crucial to ensuring
a level playing field for all market players on the freshly liberalised and partially privatised markets of public
utilities (energy, electronic communications or postal services). In the markets where the state-controlled
providers remained present, yet lost their formal monopoly and became exposed to competition of private
actors, NRAs became particularly needed to demonstrate the separation of regulatory functions from the
management of public providers. Gradually, the NRAs became the national administrative structures
subject to the most detailed EU regulation pertaining to their organisational setup and autonomy. No other
national institutions attracted comparable attention of EU legislation. Figure 3 shows the evolution of EU
standards on NRA autonomy in different policy areas. It also provides a catalogue of the NRAs covered
by the EU acquis.
20 GOV/SIGMA(2021)3
Unclassified
Figure 3. Expansion of EU law standards on organisation and autonomy of national regulatory authorities
Source: SIGMA analysis of EU legislation.
The development of EU standards has accelerated in recent years, with new EU standards established for
the national competition authorities and improvements to the existing regulation for the electronic
communications and audio-visual media regulators. However, the EU model of NRAs is relatively young,
compared to developments at the national level. The emergence of independent administrative bodies in
Western Europe can be traced back to the 1970s (Majone, 1997[40]) (Scott, 2014[41]). Since the late 1980s,
they have become widespread across Europe (Gilardi, 2006[42]). The first few EU standards emerged only
a decade later. Thus, the EU acquis cannot be perceived as a vehicle that introduced the concept of NRAs
to Europe. It was a factor contributing to dissemination, to cementing and achieving some degree of
consistency in perception of this institution across Europe. Furthermore, the formalisation of EU standards
for NRAs’ independence was often the culmination of a long-term process involving multiple actors and
actions — lobbying efforts of networks of national regulators, policy documents and soft law acts of the
European Commission and landmark rulings of the Court of Justice of the European Union (CJEU).
The content of EU standards for NRAs is not uniform. They differ particularly in terms of depth of the EU
influence on the Member States’ autonomy in shaping national administrative structures. A detailed
description of standards for each type of NRA, based on a review of the relevant EU Directives, is provided
below (Table 1).
Milestones for development of EU model of NRAs
Directive 2002/21/EC: Initial standards for NRAs in the area of electronic communications
Directives 2003/54/EC and 2003/55/EC: General obligation to establish NRA in the areas of electricity and gas market (a regulatory authority independent from industry)
Directive 2009/140/EC: Strengthening standards for NRAs in the area of electronic communications
Directive 2009/72/EC and 2009/73/EC: Elaboration of comprehensive standards for NRAs in the areas of electricity and gas market
Directive 2010/13/EU: General principle of independence of NRAs in the area of audio-visual media services (lack of provisions specifying detailed standards of independence)
Directive 2012/34/EU: Establishing comprehensive standards for NRAs in the area of railways
Directive 2018/1808: Establishing comprehensive standards for NRAs in the area of audio-visual media services
Directive 2018/1972: Upgrading the standards on independence of NRAs in the area of electronic communications
Directive 2019/1: Establishing comprehensive standards for NRAs in the area of competition protection
Directive 2019/944: Amending provisions on independence of regulatory authorities in the area of electricity
GOV/SIGMA(2021)3 21
Unclassified
Table 1. EU acquis standards of autonomy of National Regulatory Authorities
Sector and relevant acts Standards of autonomy
Energy markets Directive (EU) 2019/944 of the European Parliament and of the Council of 5 June 2019 on common rules for the internal market for electricity; Directive 2009/73/EC of the European Parliament and of the Council of 13 July 2009 concerning common rules for the internal market in natural gas
Legally distinct and functionally independent from any other public or private entity
Should not seek or take direct instructions from any government or other public or private entity
Separate annual budget allocations provided, with autonomy in the implementation of the allocated budget, and adequate human and financial resources to carry out their duties
The members of the agency’s board appointed for fixed term of five up to seven years, renewable once
Power of the government to issue general policy guidelines not related to the regulatory powers and duties
Electronic communications Directive (EU) 2018/1972 of the European Parliament and of the Council of 11 December 2018 establishing the European Electronic Communications Code
Legally distinct from, and functionally independent of, any natural or legal person providing electronic communications networks, equipment or service
Should not seek or take direct instructions from any government or other public or private entity
Only independent bodies (especially courts) should have the power to overturn or suspend their decisions
The members of management shall be appointed for a term of office of at least three years from among persons of recognised standing and professional experience, on the basis of merit, skills, knowledge and experience and following an open and transparent selection procedure. Member States shall ensure continuity of decision making
Should have separate annual budgets and have autonomy in the implementation of the allocated budget
Railways Directive 2012/34/EU establishing a single European railway area.
Independence in organisation, funding decisions, legal structure and decision making
The regulator should exist as a stand-alone authority
The managing body should be appointed under clear and transparent rules and in transparent, merit-based procedure by the government or other public authority which does not perform a supervisory role over providers of railway services
The Member States may choose one of the following solutions relating to term of office of the heads of railway sector regulators: 1) appointment for fixed and renewable term; or 2) permanent appointment with dismissal possible solely on disciplinary ground
Audio-visual media services Directive 2010/13/EU on the co-ordination of certain provisions laid down by law, regulation or administrative action in Member States concerning the provision of audio-visual media services (significantly revised in November 2018)
They should remain legally distinct from the government and functionally independent of their respective governments and of any other public or private body
They should be able to exercise their powers impartially and transparently;
They should not seek or take instructions from any other body in relation to the exercise of the tasks assigned to them under national law implementing Union law
Adequate financial (separate budget) and human resources and enforcement should be secured
Managing bodies should be appointed in transparent and non-discriminatory procedure and the members of those bodies could be dismissed only if they no longer fulfil the conditions required for their functions
22 GOV/SIGMA(2021)3
Unclassified
Competition protection Directive (EU) 2019/1 of the European Parliament and of the Council of 11 December 2018 to empower the competition authorities of the Member States to be more effective enforcers and to ensure the proper functioning of the internal market
Member States should guarantee that the national competition authorities perform their duties and exercise their powers impartially and in the interests of the effective and uniform application of those provisions, subject to proportionate accountability requirements
They should not seek nor take any instructions from government or any other public or private entity when carrying out their duties and exercising their powers
However, the government has right to issue general policy rules that are not related to sector inquiries or specific enforcement proceedings
Persons performing decision-making powers in the national competition authorities could be dismissed only if they no longer fulfil the conditions required for the performance of their duties or if they have been found guilty of serious misconduct under national law
Member States shall ensure that the members of the decision-making body of national administrative competition authorities are selected, recruited or appointed according to clear and transparent procedures laid down in advance in national law
National competition authorities shall enjoy independence in the spending of the allocated budget for the purpose of carrying out their duties
Postal services Directive 97/67/EC of the European Parliament and of the Council of 15 December 1997 on common rules for the development of the internal market of Community postal services and the improvement of quality of service
Regulatory authority should be legally separate from and operationally independent of the postal operators, i.e. within the government, regulatory responsibilities are separated structurally from the responsibility for exercising any property rights in the incumbent postal operator (ownership unbundling)
Source: SIGMA analysis of the EU legislation
Irrespective of sector-specific differences, there are some common elements in the EU standards for NRA
autonomy. First of all, the overall objective is to shield regulators from undue influence of both elected
politicians and markets on decisions made within the regulatory remit established by the EU law. NRAs
should be protected from derailing regulatory decisions either to pursue the short-term electoral goals of
political actors or to favour market players. The main instrument to realise this objective, established in EU
law, is the principle of “functional independence”, recurring in most of the Directives regulating the status
of the NRAs. This notion is key to understanding the scope of the autonomy of the NRAs required by the
EU acquis. While none of the Directives contains a definition of “functional independence”, the focus is on
ensuring that the regulatory functions of the NRAs are performed “autonomously, without any direct
possibilities of oversight by national political principals” (De Somer, 2018, p. 585[43]). In other words,
“functional independence” is about the power to exert regulatory functions, established by national law in
line with the EU acquis, with no ex ante or ex post control by elected politicians, either in the form of
instructions on the content of individual regulatory decisions or reviewing (repealing or amending) them.
Various institutional measures are prescribed to create favourable conditions for this functional
independence and reduce the opportunities for indirect external influence on regulatory decisions. The
formulation of these additional safeguards differs across Directives, but generally they concentrate on:
legal status
appointment and dismissal of the NRAs’ senior management
GOV/SIGMA(2021)3 23
Unclassified
resources.
In many cases, the standards established by the EU legislation are of a general nature and do not provide
detailed guidelines on resolving particular governance issues. This requires a case-by-case approach,
while analysing for example the scope of possible interference by the government in the management of
the regulatory authorities. A number of problematic cases have been subject to review by the CJEU, which
provided further instructions on the interpretation of the EU acquis guaranteeing the independence of
national regulators (Box 3). The Court did not expand the scope of NRAs’ autonomy beyond the standards
established by the Directives but translated some of the general rules into more specific guidelines. The
most important general conclusion that could be drawn from these rulings is the prioritisation of the
functional independence of the NRAs, i.e. ensuring that they are capable of implementing regulatory
measures stemming from EU law without undue external influence.
Box 3. CJEU case-law on independence of national regulatory authorities
The case C-424/07 Commission v. Federal Republic of Germany establishes that the national legislation cannot reduce the regulatory remit established by EU legislation for the national telecommunications regulator. Furthermore, national law cannot impose hierarchy within the catalogue of regulatory objectives established by the EU legislation as guiding principles for the national regulator. Striking a fair balance between these objectives is the responsibility of the national regulator. In the case C-560/15 Europa Way and Persidera, the CJEU declared violation of the EU law resulting from the adoption of the law cancelling the on-going selection procedure for the allocation of radio frequencies. The Court noted that the law was an example of illegitimate interference of a political body in administrative decision-making within the regulatory remit of the NRAs. Interference should be confined to review of the regulator’s decisions by the court or other independent body. The Parliament has no power to interfere, even through legislative measures, in pending administrative proceedings before the regulatory body implementing its functions established by the EU legislation. Financial autonomy of regulators was the subject of the case C-240/15 Autorità per le Garanzie nelle Comunicazioni. Italian law had reduced annual appropriations and year-to-year budget increases applicable to a large group of public authorities, including the electronic communications regulatory body. In this case, the CJEU did not find this intervention in financial management of the regulator illegal, as it was not demonstrated that these budgetary restrictions undermined the regulator’s capacity to perform its core functions or were designed in a manner that was discriminatory to the NRA. Crucial issues of agency governance were considered by the CJEU in the ruling on case C-424/15 Xabier Ormaetxea Garai, Bernardo Lorenzo Almendros v. Administración del Estado. The Court analysed the transitional provisions of the law that consolidated the Spanish sectoral regulators into a single National Authority for Markets and Competition (see above). This law terminated the mandate of the members of the management of the amalgamated bodies, who were appointed for a fixed term. While the CJEU did not contest the amalgamation itself, it pointed out that the sole ground of restructuring the agency does not justify premature dismissal. The Court advised that, according to the relevant provisions of the EU law, members of the management of an NRA operating in the area of electronic communications may be dismissed only if they no longer fulfil the conditions required for the performance of their duties, which are laid down in advance in national law. This prerequisite was not fulfilled in the Spanish case, as the early dismissal resulted from other grounds, namely the reorganisation of the institution. The position of the CJEU does not impede reorganisation as such, but requires the Member States to ensure continuation of the mandate of the management of the body subject to restructuring. For example, they could be transferred to the management of the body emerging from the reorganisation. Source: SIGMA review of the CJEU case law.
24 GOV/SIGMA(2021)3
Unclassified
The last case mentioned above is particularly interesting in the context of organisational reform of
agencies, as it establishes special requirements pertaining to the restructuring of NRAs. The CJEU
emphasised that reorganisation should not be misused as a method for premature dismissal of members
of the management of NRAs and their mandate should be “transferred” to the newly-created body.
Importantly, however, the Court also confirmed that regulatory functions established by the EU law in
various domains could be organised under multi-sectoral regulatory agencies. There are no obstacles to
seeking efficiency gains in the regulatory sphere through the consolidation of NRAs.
In practice, this trend towards the amalgamation of regulators is already well-established across the EU.
The table below presents an overview of multi-sectoral regulators in the EU, showing which of the
regulatory domains governed by the EU legislation are grouped under multi-sectoral regulators. For the
purpose of this review, we selected regulatory authorities covering at least three EU regulatory domains.
Table 2. Multi-sector regulators in EU member countries
Audio-visual
media services
Electronic
communications
Energy Railway Postal services Competition
protection
Croatia ● ● ● ●
Estonia5 ● ● ● ●
Germany ● ● ● ●
Lithuania ● ● ●
Luxembourg ● ● ● ●
Netherlands ● ● ● ● ●
Slovenia ● ● ● ●
Spain ● ● ● ● ●
Source: SIGMA review of national legislation.
Notes: NRAs may have other sectors under their purview. For example, economic regulation of the water sector (sometimes grouped with
energy regulation) and air transport regulation (sometimes grouped with railway regulation) (Casullo, Durand and Cavassini, 2019[44])
While member countries are free to adjust the rules established by the EU Directives to their national
administrative tradition and legal system, in the case of energy market regulators (electricity and gas), the
European Commission issued an Interpretative Note (European Commission, 2010[45]) providing additional
guidelines on the institutional setup for these bodies (Table 3). It should be emphasised, however, that
these guidelines are not formally binding and they do not provide the only possible interpretation of the
rules established by the Directives.
5 This relates to the Estonian Competition Authority. It should be noted that responsibility for regulation of audio-visual
media services and electronic communications is merged under another multi-sectoral regulator (Consumer Protection
and Technical Regulatory Authority) that is also responsible for consumer protection.
GOV/SIGMA(2021)3 25
Unclassified
Table 3. European Commission guidelines on interpretation of independence standards for electricity and gas market national regulators
Aspect of independence EC guidelines
Legally distinct status The NRA must be created as a separate and distinct organisation from any Ministry or other government body. The NRA can no longer be part of a Ministry. This includes a recommendation against sharing personnel and offices with a ministry.
Functional (decision-making) autonomy
If the NRA is to draft a work programme for the coming year(s), it should be able do so autonomously, i.e. without the need for the approval or consent of public authorities or any other third parties.
Decisions by the NRA cannot be subject to review, suspension or veto by the government or the ministry (judicial review available)
Financial autonomy The NRA’s budget may be part of the state budget.
The NRA may neither seek nor receive any instruction on its budget spending.
Power of the government to issue general policy guidelines
The independence of the NRA does not deprive the government of the capacity to set national (energy) policy. It may be within the government’s competency to determine the policy framework within which the NRA must operate, e.g. concerning security of supply, renewables or energy efficiency targets. However, general energy policy guidelines issued by the government must not encroach on the NRA’s independence and autonomy.
Source: Based on European Commission 2010
The provisions of the Directives and the CJEU’s jurisprudence and soft law standards define the EU
standards for the autonomy of the NRAs. However, in order to effectively address misunderstandings and
misinterpretations that are common in practice in the Western Balkans and the European Neighbourhood,
more practical guidance is needed for designing the organisational setup of the NRAs in line with the EU
requirements. The table below presents tailored guidelines, explaining what is and what is not required by
the EU acquis in each dimension of the agencies’ autonomy, as described in the first chapter (Figure 1).
Table 4. Clarifying the NRA’s autonomy. What is and what is not required by EU law
Autonomy dimensions
EU law requirements for autonomy for and legitimate restrictions of NRAs
Legal autonomy - requirements
The NRA should have the status of a standalone organisation, not an internal unit of a ministry. The following attributes are essential to ensuring this status: a) capacity to act as a standalone administrative body in administrative procedure; b) authorisation to enter into contractual relations and act as a contracting authority in the public procurement system; c) separate budget from the ministry’s budget; d) capacity of the senior management to execute basic managerial powers relating to organisational, financial or human resource management (HRM) matters; e) capacity of the senior management to configure internal structures of the NRA in line with the general rules established by law.
Legal autonomy - legitimate restrictions
Legally distinct status does not require the NRA to have the status of a separate legal person under private law, (it may operate within the unique legal personality of the state). Legally distinct status refers to the NRA having some attributes of legal and organisational identity that ensure that it operates as a standalone organisation.
Merging regulators regulated by different Directives into single multi-sectoral regulator is fully legitimate, as long as all the required guarantees of autonomy are secured for this new body. In cases of this kind of transformation, the mandate of the management of the amalgamated bodies should not be automatically terminated, but rather “transferred” to the new body.
Policy and The objectives and powers of the NRA should be established by the national law in line with the EU law.
26 GOV/SIGMA(2021)3
Unclassified
interventional autonomy - requirements
The national legislation cannot reduce the regulatory powers allocated by the EU law or determine the hierarchy of regulatory objectives.
The government or legislature cannot issue instructions or guidelines about the manner of executing regulatory powers (e.g. ordering to lower tariffs or block specific transactions) or interfere in the pending administrative proceedings, e.g. by terminating procedure or ordering to undertake specific investigatory activities.
The government (or legislature) should abstain from imposing any form of sanctions associated with their assessment of the regulatory decisions of the NRAs, especially in the form of dismissal of their management, budget cutting or public reprimand.
Decisions (administrative acts) undertaken by the NRA within its regulatory remit can be subject to review only by the court. The ministry or any other government body cannot act as an appeal body.
Policy and interventional autonomy – legitimate restrictions
The government (portfolio ministry) may issue general policy guidelines reflecting the government’s policy for the relevant sector. These policy guidelines should not take the form of instructions regarding the resolution of particular cases, but may include recommendations on e.g. tackling specific types of irregularities and distortions in the market, enhancing competition on the market or protection of consumers’ rights, contributing to the objective policy targets. The NRA may also be asked to provide input to the policy formulation process, by developing proposals or sharing comments on the portfolio ministry’s proposals.
The EU acquis does not regulate horizontal standards for performance management and reporting. Member States are able but not obliged to ask regulators to prepare annual plans and reports, specifying objectives and targets in a format defined by legislation or government acts. NRAs might be required to discuss the plans and reports with the portfolio ministry, though no formal approval of the ministry should be required. The performance of the NRA may be discussed regularly with the portfolio ministry, but without formal scoring or appraisal.
As the NRAs are required to act in a transparent manner, the government (portfolio ministry) has the right to obtain information and explanations about the NRA’s activities and position in the regulated sector, observing special rules relating to the protection of commercial secrets and other types of confidential information.
The legislature may exert classical instruments of parliamentary oversight towards NRAs, including interpellations, requests for information, hearings of the senior management, discussing annual reports or requesting the supreme audit institutions to conduct audits of the respective NRA.
Structural autonomy - requirements
The members of the management of the NRA should be appointed in a transparent, non-discriminatory and merit-based procedure for a fixed or open-ended mandate. In the case of electronic communications and energy markets regulators, a fixed-term appointment is required, respectively for at least three up to seven and five up to seven years.
Irrespective of the term of appointment, early dismissal of senior management should be restricted to situations when they longer fulfil the conditions required for the performance of their duties. These conditions should be established in advance by the national law and may specifically refer to serious misconduct, misbehaviour or conflict of interest, undermining the neutrality and objectivity of the NRA. Discontentment with regulatory decisions made by the NRA or generally formulated unsatisfactory performance cannot serve as grounds for early dismissal. Dismissal should be justified.
Structural autonomy – legitimate restrictions
No specific structure of management is required. In particular, both monocratic and collegial management is allowed. There is no requirement for dual management structure (management + supervisory board).
There is no requirement for involvement of the parliament in appointment or dismissal of the senior management. Decisions may be taken by the government as a whole or by the portfolio ministry. In the case of railway market regulators, appointments must be made by a public authority that does not perform a supervisory role over providers of railway services.
Financial autonomy - requirement
The NRA can be financed from the state budget. Alternative funding options may include fees collected from the market operators. There should be a separate allocation for the NRA – its budget cannot constitute part of the portfolio ministry’s budget.
GOV/SIGMA(2021)3 27
Unclassified
The budget should be sufficient for effective performance of the regulatory functions, although there are no objective criteria established by the EU law to assess the adequacy of the budgetary allocation. The allocated budget should be managed (executed) by the NRA autonomously, observing the general rules of public financial management pertaining to the release of public funds.
Financial autonomy – legitimate restrictions
The budget of the NRA can be adopted in the standard government budgetary procedure, i.e. negotiating the allocation with the ministry of finance, followed by approval of the budgetary proposal by the government and adoption by the parliament. The budgetary allocation may take the form of a lump sum (total allocation) or a more detailed breakdown (e.g. staff, capital expenditure).
Budgetary cuts (e.g. on a year-to-year basis) are generally allowed, but the context is important for the assessment of their compliance with the EU law. If the cuts are based on objective grounds (e.g. budgetary crisis) and apply to all or most of the public authorities, they could only be challenged if they seriously hindered the NRAs’ capacity to perform its basic functions. In such a case, the NRA should provide clear evidence of the significant and detrimental impact of the cuts. Income generated by the regulator (e.g. from fees or fines) may be transferred to the state budget or retained by the NRA proportionately and under conditions specified by law.
Managerial autonomy - requirements
Sufficient staffing should be secured, although there are no objective criteria established by the EU law to assess the adequacy of the staff capacity of the NRA.
The staff and the allocated budget should be managed by the NRA autonomously. This appears to exclude the interference of any external body on individual decisions for recruitment, allocation, promotion or termination of employment of NRA staff.
Once the NRA’s budget is adopted, the spending decisions within this budget, in principle, should be made without a special external approval process enabling any external body to decide arbitrarily whether specific spending should be allowed or not.
Managerial autonomy – legitimate restrictions
The staff of the NRA can be part of the civil service system of the government. The rules regarding recruitment, categorisation of staff, allocation, promotion, performance appraisal or termination of employment may apply fully to the personnel of the NRAs, as long as they do not envisage external interference in decisions relating to individual staff members, e.g. in the form of appointment of the heads or staff of the NRA’s internal units by the portfolio ministry. Sublegal acts issued by the government, e.g. setting detailed rules on performance appraisal or disciplinary proceedings may apply as long as they do not imply external involvement in appraisal or disciplinary procedure pertaining to individual staff members.
Legislation on salaries can apply fully, including the salary scales determined by law or by the decision of the government based on the delegation in the law. Individual application of these rules should remain the decision of the NRA’s management.
The general rules on expenditure control within the government may apply to the NRAs fully. Depending on the system of expenditure controls adopted in the relevant country (see: International Monetary Fund 2016), this may include apportionments by the ministry of finance (granting spending authority for specific periods), centralised payments, monitoring budget execution by the ministry of finance or system of ex ante approvals of the ministry of finance for specific types of transactions.
The NRA may participate in a cross-government shared services system (provision of back-office services by a single body for multiple government institutions), though on a voluntary rather than a mandatory basis, especially if this covers such issues as centralised procurement or recruitment.
The legality and integrity of financial, HRM and operational management could be subject to inspections conducted by the portfolio ministry, ministry of finance and other competent government bodies. This does restrict the inspection powers of independent accountability bodies, such as supreme audit institutions, anti-corruption agencies or ombudsmen.
Source: SIGMA analysis of the EU acquis.
The EU concept of functionally independent regulators corresponds well with the guidelines produced by
OECD in the Recommendation of the Council on Regulatory Policy and Governance and Best Practice
28 GOV/SIGMA(2021)3
Unclassified
Principles for Regulatory Policy: The Governance of Regulators (OECD, 2014[46]), followed by more recent
Practical Guidance against Undue Influence (OECD, 2017[47]). The OECD also clearly locates regulators
within the public administration where they function with sufficient autonomy within the powers delegated
by the legislature and national policy frameworks. The 2014 Best Practice Principles for Regulatory Policy
explicitly provide room for government (ministerial) stewardship of regulators that should rely on the powers
clearly defined in the legislation. This is intended to uphold alignment "between the long-term policy goals
of the regulator and the broad, strategic national priorities as set by elected representatives in the
executive, congress or parliament” (OECD, 2016, p. 4[48]) (OECD, 2014[46]).
The OECD instruments (see Box 4, and in particular Principle 7 of the Recommendation of the Council on
Regulatory Policy and Governance) reaffirm the importance of transparent appointment procedures for the
senior management of the regulator, emphasise the importance of the stability of mandates and for
restricting the grounds for early dismissal to specific types of misconduct or incapacity. In financial matters,
OECD emphasises the benefits of multi-annual budgetary allocations, in order to reduce the risk of undue
pressure during annual budgetary negotiations. The need for securing spending autonomy is another
dimension of financial autonomy (OECD, 2017[47]) (OECD, 2014[46]).
Box 4. Selected OECD guidance for independence of regulators
Legal/functional autonomy
The mandate of the regulators should be established in the legislation.
They should regularly report to the executive and/or parliament on its performance and operate in line with long-term national policy for the relevant sector.
Governments and/or the legislator should monitor and review periodically that the system of regulation is working as intended under the legislation. In order to facilitate such reviews, the regulator should develop a comprehensive and meaningful set of performance indicators.
While performing their regulatory functions autonomously, they should also be required to co-operate with other government bodies and provide advice in policy making processes.
Appeals against the decisions of the regulators should be considered by an independent body located outside the government.
Structural autonomy
Management should be appointed by transparent procedure with selection criteria known in advance.
The number of terms of appointment to the management of the regulators should be limited.
Special conflict of interest rules should apply to the members of management, including conflict of interest registers, declaration of assets/shares/interests or duty to publish justifications of the key decisions.
The grounds for dismissal should be established in the law and limited to serious cases of misbehaviour.
Financial/managerial autonomy
The budget can be decided on a multi-year basis (e.g. for three years) and should be spent with appropriate and accountable autonomy with the general rules of public spending and procurement, as well as auditing, fully applicable.
GOV/SIGMA(2021)3 29
Unclassified
Regulators should not set the level of their cost recovery fees, or the scope of activities that incur fees, without arm’s-length oversight.
Source: Based on: OECD Best Practice Principles for Regulatory Policy; Creating a Culture of Independence, Practical Guidance against
Undue Influence; 2012 Recommendation of the Council on Regulatory Policy and Governance
(OECD, 2014[46]) (OECD, 2017[47]) (Casullo, Durand and Cavassini, 2019[49])
The box below provides brief descriptions of multi-sector NRAs in Ireland and Latvia, demonstrating the
national interpretations of the EU’s principles of NRA independence. In some aspects (e.g. funding
regime), they go beyond the minimum required by the EU acquis.
Box 5. NRA independence in action (Latvia, Ireland)
Two performance assessment peer reviews produced by the OECD Network of Economic Regulators provide useful insight on specific ways to safeguard functional autonomy. The Latvian Public Utilities Commission (PUC) is a multi-sector regulator responsible for electronic communications, energy, postal services, water management and waste disposal. The law defines the PUC as an autonomous public institution, with institutional and functional independence and an independent balance sheet and account with the Treasury. The PUC is financed through fees from the regulated sectors established by law and calculated as a percentage of the net turnover of public utilities, and the Ministry of Finance incorporates the budget within the state budget proposal that is approved by Parliament. The PUC also has other arrangements in place designed to foster its independence. For example, it does not receive guidance from the Government related to regulatory decisions, and board members are subject to cooling-off periods after their term ends. Moreover, the PUC remains accountable to multiple public bodies. It is legally bound to report yearly to Parliament, and it is subject to the control of the State Audit Office. The PUC’s accountability to Parliament involves the presentation of an annual action plan, including strategic and operational objectives. This is complemented with the obligation to present to the legislature a report on its activities and an audited financial statement annually. In addition, the PUC has also been invited to provide evidence in the parliamentary committees on a number of issues related to the sectors regulated by the PUC (OECD, 2021[50])It is also subject to the control of the State Audit Office. Beyond providing an advisory function and issuing formal opinions to Ministries, the PUC is proactive in co-ordinating and exchanging information with ministries and other key institutions. Regular informal exchanges at the technical and political levels with sector ministries and other key institutions complement formal processes and co-ordination agreements. There are regular meetings between, for example, the Minister of Economy, State Secretaries and the PUC Board. In addition, staff from key ministries and the State Secretary meet quarterly with the PUC Board. The ability of regulated entities or members of the public to appeal decisions of the regulator is another important control on regulatory decision making. Consumers and regulated entities can appeal an administrative act or action of the PUC to an Administrative Regional Court, which adjudicates the matter as a court of first instance (OECD, 2016[51]). Ireland’s Commission for the Regulation of Utilities (CRU) is a multi-sector regulator with responsibilities for economic regulation of the electricity, gas and water sectors as well as for energy safety. Its founding legislation states that the CRU “shall be independent in the performance of its functions” (Electricity Regulation Act as amended, 1999). In accordance with its formal independence, the CRU maintains independence over its budget, which is funded entirely through levies. The Government and legislature do not review or approve the levies, and the levies are not part of the government budget. The CRU’s legal and regulatory framework specifies certain arrangements designed to maintain the CRU’s independence. For example, CRU Commissioners are subject to a one year cooling-off period after their terms.
30 GOV/SIGMA(2021)3
Unclassified
The CRU is under the aegis of two ministries, the Department of Communications, Climate Action and Environment (DCCAE) and the Department of Housing, Planning and Local Government (DHPLG). The CRU provides annual work plans to the two Ministers. The regulator also submits an annual report to the Irish Parliament via the DCCAE, which presents the materials to Parliament. This channel of communication reflects the requirements of the legislation for the communication of such documents to Parliament rather than any formal approval or appraisal of the documents by the Department. The CRU provides both informal and formal inputs to the DCCAE and DHPLG on policy formulation. Formal input takes the form of participation in consultation processes and provision of expertise. In some instances, such as when providing feedback to EU legislation, the DCCAE relies on CRU technical expertise. In the water sector, the DHPLG has called on the CRU frequently to provide advice on matters pertaining to CRU duties. Indeed, according to Section 40 of the Water Services Act (2) 2013, the CRU may also advise the Minister on the development and delivery of water services. Informal input takes the form of information-sharing through meetings and personal connections. CRU senior management meets with Ministers or senior management in the respective Departments on a regular basis to share information about activities and sector issues. The CRU also provides informal feedback to DCCAE on legislation. Given the small size of the administration, informal exchanges between staff of the CRU, the DCCAE and the DHPLG are common and are seen to aid effective communications. While these informal interactions keep communication channels open, the OECD peer review recommended that CRU build more transparency into information-sharing and co-ordination channels with the executive and maintain a more structured and predictable relationship with Parliament in order to enhance the CRU’s accountability. Regulated entities can appeal the CRU’s regulatory policy decisions through judicial review in the High Court or the Commercial Court. Minister for DCCAE
(OECD, 2018[52]) (OECD, 2021[50]).
Expanding the standards of autonomy to non-regulatory bodies
EU regulations on the institutional setup of agencies are focused primarily on NRAs, but also cover some
aspects of non-regulatory bodies. The notion of independence recurs in different contexts, also with regard
to other types of agencies. In some cases, the guarantees of autonomy are of similar scope. For others,
they are limited to functional autonomy in performing core functions. Even for experts, the fragmentation
of these standards and lack of uniform definitions of key concepts make it difficult to understand the exact
requirements set out by the EU. The European Commission therefore produced the “Guide to the main
administrative structures required for implementing the acquis” for EU candidate countries and potential
candidates in 2013. However, this guide was never published and has not been updated. SIGMA therefore
conducted a thorough review of these standards to provide updated and detailed guidance.
Extensive autonomy requirements have been established for the national data protection authorities
(Regulation (EU) 2016/679). In addition to ensuring their complete independence and freedom from
external influence in performing their tasks, EU member countries are required to guarantee:
Sufficient human, technical and financial resources, premises and infrastructure.
Autonomy in recruiting and managing staff.
Separate budget allocation.
Members of the management of national data protection authorities should be appointed through a
transparent procedure by the parliament, the government, the head of the state or a designated
independent body (e.g. bodies of the judicial branch). Early dismissal of the senior management should be
possible solely in cases of serious misconduct or situations where the respective member no longer fulfils
the conditions required for the performance of their duties.
GOV/SIGMA(2021)3 31
Unclassified
Some guarantees of the functional autonomy of national equality bodies have been established by a
series of Directives (Directives: 2000/43/EC, 2004/113/EC, 2006/54/EC, and 2010/41/EU). Their mission
is to tackle various forms of discrimination on grounds of gender and racial or ethnic origin. Equality bodies
are ombudsman-type institutions protecting and promoting equal treatment. They may operate either as
standalone, specialised equality bodies, or their functions might be performed by national human rights
institutions. There are several examples of these functions being performed by the ombudsperson
institution (Bosnia and Herzegovina, Czech Republic, France, Georgia, Greece, Kosovo, Latvia,
Montenegro, and Poland). Regardless of the institutional model, EU member countries are obliged to
enable them to provide independent assistance to victims of discrimination, conduct independent surveys
and publish independent reports.
Another group of non-regulatory bodies enjoying special guarantees of autonomy are safety investigation
authorities in air, maritime and railway transport (Regulation (EU) No 996/2010, Directive (EU) 2016/798
and 2009/18/EC). The civil aviation safety investigation authority must be functionally independent in
particular of aviation authorities responsible for airworthiness, certification, flight operation, maintenance,
licensing, and air traffic control or aerodrome operation. They should also be functionally independent from
any other party or entity, the interests or missions of which could conflict with the task entrusted to the
safety investigation authority or influence its objectivity. It is not explicitly determined whether the civil
aviation safety investigation authority could remain within the organisational structure of the respective
ministry, but in such a case, it should definitely enjoy greater guarantees of autonomy than a typical
organisational unit of the ministry. Safety investigation authorities in maritime transport and railways should
be provided with independence in their organisation, legal structure and decision making from any party
whose interests could conflict with the task entrusted to them. This does not imply, however, the need to
create an ‘independent enclave’ outside the public administration, but rather to secure this authority from
interference in performing impartial investigations. In the case of national safety authorities for railways,
the Directive explicitly states that this body can be situated within the national ministry responsible for
transport matters. This possibility also applies to other safety investigation authorities, as their location
within the ministry does not pose inherent conflict with their functions.
Specific standards of independence are also set for quasi-judicial, dispute resolution bodies, including
procurement review bodies (Directive 89/665/EEC). If EU member countries decide to establish
procurement review bodies of non-judicial character, they are required to ensure that members of such
bodies are appointed and dismissed under the same conditions as members of the judiciary, as regards
the authority responsible for their appointment, their period of office, and their removal. Further, at least
the President of this independent body shall have the same legal and professional qualifications as
members of the judiciary.
The autonomy of national statistical institutes (NSIs) is subject to EU regulation (Regulation (EC)
223/2009). In this case, the EU law introduces the concept of ‘professional independence’ of officials
performing the NSIs’ tasks, defined as the power to autonomously develop, produce and disseminate
statistics, with any instructions from government or other bodies, regarding performance of this core
mission, precluded. The content of this principle is the same as “functional independence” used in the
context of the NRAs. In both cases, the EU law focuses on ensuring autonomous performance of the core
functions of relevant bodies. Similarly to the NRAs, the Regulation (EC) 223/2009 contains some
subsidiary requirements on other aspects of the NSIs’ autonomy, such as autonomy in internal
management, or the principle of transparent and merit-based appointment of the heads of the NSIs. As
regards grounds for dismissal of the heads of the NSIs, there is a general rule that it should “not
compromise professional independence”. This allows for greater autonomy of the government to remove
the heads of the NSIs (e.g. based on negative performance assessment), although dismissal as a sanction
for refusing to follow the government’s instructions on performing core functions would still be illegitimate.
Finally, in several cases the EU acquis uses the attribute of “independence” of some bodies, but in the
specific context of separating them from other institutions and parties in order to prevent conflict of interest,
32 GOV/SIGMA(2021)3
Unclassified
not with the intention of creating organisations enjoying special status within the government
administration. These bodies remain classic government agencies (or ministries), but should be
institutionally separated from the bodies they oversee. This applies to the following institutions:
National accreditation bodies should be independent from the conformity assessment bodies they
assess (Regulation (EC) 765/2008).
Market surveillance authorities should operate independently from the market operators they
control (Regulation (EC) 765/2008).
Certification bodies should be independent from paying agencies managing the payments from
the Common Agricultural Policy, which are subject to auditing by the certification agencies
(Regulation (EU) 1306/2013).
National enforcement bodies protecting the rights of passengers when travelling by sea and inland
waterway should be independent from market operators (Regulation (EU) 1177/2010).
Regulatory authorities in the field of radioactive waste management should be separate from any
body concerned with the promotion or utilisation of nuclear energy or radioactive material (Council
Directive 2011/70/Euratom).
Dispute settlement bodies for disputes on access to transport networks and to storage sites for the
purposes of geological storage of produced and captured carbon dioxide should be independent
from the parties of such disputes (Directive 2009/31/EC).
Summary – EU standards for national agencies
Generally, national governments can decide on the organisation of the public administration, yet the scope
of EU acquis regulating these matters has gradually expanded to support effective implementation of the
EU legislation by the national administration. The goal of EU legislation is to shield the respective national
authorities from undue influence and pressure from the markets, interest groups and elected politicians.
NRAs are of particular interest to the EU, resulting in the most extensive regulation.
Nonetheless, the NRAs do not constitute a “fourth branch of government”. The central attribute required
by the EU legislation is functional autonomy, i.e. formally guaranteed powers to apply their regulatory
powers with no external influence, except for judicial review of regulatory decisions. All other standards
relating to legal status, composition of management, financial or HRM matters differentiate the status of
NRAs to only a limited extent from other public agencies. As such, NRAs remain firmly located within the
executive branch of the state.
The EU did not invent the notion that elected politicians should not interfere in individual proceedings or
decisions, or that civil servants should be bound by law, not the will of political leaders. However, EU law
is more explicit in curbing ministerial power over administrative acts of the NRAs. However, when
translated to the EU Enlargement (Western Balkan) and neighbourhood context these realities get lost in
translation.
GOV/SIGMA(2021)3 33
Unclassified
Ministerial management and oversight tools to balance agency autonomy and
control
All governments search for the same ideal in managing agencies – a regulatory and institutional framework
that strikes the perfect balance between agency autonomy and control. A perfectly-balanced framework
would grant extensive autonomy to agencies, but at the same time ensure oversight of the results achieved
and allow for adequate control and risk management. Achieving this equilibrium means avoiding the
micromanagement of agencies on the one hand and a governance vacuum on the other.
Micromanagement undermines the very purpose of creating agencies – if the agency is managed in the
same way as an organisational unit in a ministry, what is the added value of its formally distinct
organisational status? A governance vacuum, characterised by a lack of clearly-defined expectations
towards agencies, a lack of results-oriented performance management schemes and random interventions
in their operations, is problematic because the portfolio ministry remains ultimately accountable for the
performance of the whole policy area, including the tasks transferred to the agencies.
Table 5. In search of equilibrium: calibrating ministerial management of agencies
Limited influence of the ministry (government) on governance of agency, e.g. senior management appointed by the parliament, work plan and budget adopted solely by the parliament;
Lack of clear reporting obligations towards the ministry;
Lack of supervisory powers of the ministry, e.g. right to issue guidelines;
The agency articulates government policy for the sector
Limited interference of the ministry in day-to-day management, e.g. confined to ex ante approvals of some activities, ex post financial, compliance and performance auditing;
Ministry empowered to issue general policy guidelines;
Objectives and targets negotiated and agreed between ministry and agency
Unrestricted power of the ministry to appoint/dismiss the senior management
Extensive influence of the ministry on day-to-day management, e.g. staff management, internal structures, ex ante approvals required for most of the operations (e.g. spending decisions);
Unrestricted powers of the ministry to issue binding instructions and guidelines on all issues;
Objectives and targets imposed compulsorily by the ministry
It is easy to agree on a balanced governance approach as an overarching goal, but harder to translate this
objective into specific guidelines. A useful first step is to establish basic rules describing the relationships
between a ministry and the agencies operating in its domain. These rules should ensure that agencies are
accountable for implementing policies shaped by the government and clarify the degree of agency
autonomy necessary by making a clear distinction between legitimate and excessive forms of ministerial
interference for different types of agency with different levels of autonomy. These rules should be
established in the relevant legislation. It is recommended that, where the legal framework is not robust,
governance models and organisational arrangements are developed when a new agency is created. Once
III. International experience and tools for
agency governance
34 GOV/SIGMA(2021)3
Unclassified
autonomy is granted, in particular in a transition environment, agencies may use their relationships with
stakeholders and media to resist changes they dislike. (Beblavý, 2002, p. 32[53])
Ministerial management represents only one layer of the oversight and accountability applicable to
agencies. Properly designed, a comprehensive model for agency accountability would address multiple
layers. It involves numerous actors overseeing the actions of agencies. Some of them have the power to
impose sanctions (courts, parliaments) or grant rewards (parliaments) to the agencies, others are only
empowered to request information or conduct investigations (independent oversight bodies, general
public).
Figure 4. Layers of agency accountability
General public: Requesting public
information; providing customer
feedback on performance
Independent oversight bodies:
conducting investigations on alleged
violations of human rights by the
agency (Ombudsperson); financial,
compliance and performance external
auditing (supreme audit institutions)
Courts: Reviewing the administrative
acts of the agency
Parliament: Requesting information
and documents; arranging hearings
regarding the agency’s activities; in
some cases, involvement in the
appointment of senior management;
budgetary oversight by parliamentary
finance committee
The degree of autonomy from the ministry remains the central criterion distinguishing various types of
agencies. This is demonstrated in international practice. In some countries, an official typology of agencies
is established by organic laws on public administration. In other cases, there is no official typology, but
different types of agencies are commonly recognised. The table below provides examples of typologies of
administrative agencies in selected countries representing various European administrative traditions.
General public
Courts
Independent oversight bodies (SAIs,
ombudsperson)
Parliament
Government (via portfolio ministry)
Agency
GOV/SIGMA(2021)3 35
Unclassified
Table 6. Typologies of public agencies in selected European countries
Country Types
Bulgaria
State agencies – established by law or decree of Council of Ministers; budgetary legal entity directly subordinated to the Council of Ministers responsible for developing and implementing policies in the areas where the ministry was not
established. Head is appointed by the Council of Ministers.
State commissions – established by law or decree of Council of Ministers; budgetary legal entity collegial body directly subordinated to the Council of Ministers responsible for licenses, permits and control. Members are appointed by the
Council of Ministers.
Executive agencies – established by law or decree of Council of Ministers; body accountable to the relevant ministry,
responsible for provision of administrative services. Head is appointed by the relevant minister
Germany
Direct administration – no legal personality, operate under legal and functional supervision of parent ministry, their
budget is part of ministry’s budget, usually governed by monocratic managing body
Indirect administration – legal personality of public law, restricted supervision of the parent ministry (legal supervision),
separate budget, more flexibility in HRM, often managed by collegial boards consisting of members appointed by
various actors
Malta
Departments – default option for performing policy implementation functions under direction of the relevant ministry
Agencies – body having a separate and distinct legal personality and capable of entering into contracts, of employing personnel, of acquiring, holding and disposing of any kind of property for the purposes of its operations, and of suing
and being sued. Government may issue instructions to ensure that agencies coordinate their activities with other agencies, departments, government entities and local councils as applicable; put into effect measures to improve the performance of agencies and the quality of the services they deliver to the public. The Government should enter into
performance agreement with each agency
Netherlands
Contract (executive) agencies – have no legal personality and are fully subordinated to the portfolio ministries;
autonomous in day-to-day management, but with limited financial autonomy
Autonomous Administrative Bodies (ZBOs) – more autonomous, most of them have legal personality of public or private
law. Ministries supervise them mainly through performance agreements
Slovenia
Bodies within ministries – created by government regulation, lack separate legal personality, operate under direct supervision of the minister, financed via financial plan adopted by the minister, minister fully responsible for their
performance
Public agencies – created by special laws, have separate legal personality, enjoy operational independence, supervised
by the relevant ministry, financed via appropriations agreed with the ministry or via fees defined in legislation
United
Kingdom
Executive agencies – remain part of the government departments, established in order to perform policy implementation
functions under departmental hierarchy, staffed by civil servants, included in the department’s budget
Non-departmental public bodies – operate at arm’s length from ministers that are responsible for their overall
performance, have separate budget, department establishes their strategic framework
Non-ministerial departments – they are similar to normal government departments in terms of functions, but are free from direct political oversight of the ministers, have separate budget, set own delivery policies, although relevant
ministry may set strategic framework
Source: Law of Bulgaria on Administration, no. 130/5.11.1998; The 2002 Law of Slovenia on Public Administration; The 2002 Law of Slovenia
on Public Agencies; The 2019 Public Administration Act of Malta; Cabinet Office 2016; Bach 2012; Yesilkagit & Van Thiel 2008.
The table shows a rather consistent approach to the design of agency architecture across countries. The
most common approach seems to be the distinction of two or three types of agencies, differentiated by the
institutional distance from the portfolio ministry, i.e. the dimensions of accountability. However, it should
be emphasised that even highly autonomous agencies remain embedded in the public administration, with
extensive oversight, co-ordination and control powers of the government, in particular the portfolio ministry.
The regulation of relations between ministries and Independent Administrative Bodies (ZBOs) in the
Netherlands is a good illustration of this model (Box 5). ZBOs represent the more autonomous type of
Dutch agencies (including regulatory authorities), but they remain firmly located within the administrative
apparatus of the Government.
36 GOV/SIGMA(2021)3
Unclassified
Box 6. Ministerial powers towards Independent Administrative Bodies (ZBOs) in the Netherlands
Governance: The portfolio ministry has the power to appoint and dismiss the members of the senior management of the ZBOs. It also decides on the salary of the senior management. Policy matters: With regard to policy issues, the ministry may set policy rules relating to performance of the agency’s core tasks. Further, it may reverse a decision made by a ZBO (with the exception of regulatory authorities, where the EU law precludes ministerial review of decisions made by the regulator). The law also provides the ministry with general competence to take whatever measures are necessary should a ZBO ‘seriously neglect its duties’. The ZBO is required to submit its annual report to the ministry. Financial management: Budgetary proposal of the ZBO should be submitted to the portfolio ministry for approval. Ex ante approvals: The ministry may decide that the following actions of the ZBO require ex ante approval of the ministry: the establishment or acquisition of an interest in a legal person; the acquisition of title to, the alienation or the encumbrance of registered property; the conclusion of credit agreements and loan agreements; the conclusion of agreements whereby the autonomous administrative authority undertakes to provide security, including security for third-party debts, or whereby it binds itself as guarantor or joint and several debtor or warrants performance by a third party; filing for bankruptcy or protection from creditors. Source: The 2006 Autonomous Administrative Authorities Framework Act
In other European countries, the portfolio ministries are provided with similar powers towards agencies
operating in their respective policy domains. The table below extracts ministerial management instruments
from the framework laws on public administration in selected European countries. It focuses on the
countries where the respective framework laws on public administration regulate this issue generally for
all agencies. It demonstrates a relatively consistent approach, with similar tools available to the portfolio
ministries in order to ensure consistent implementation of the laws and government policies by the sub-
ministerial apparatus.
Table 7. Selected European countries with framework laws – examples of ministerial management instruments
Country Ministerial management instruments
Estonia
Towards subordinated executive agencies and inspectorates:
Adopting statute;
Creating, reorganising and terminating local branches;
Appointing and dismissing heads of the institutions;
Monitoring performance;
Exercising supervisory control over the decisions in terms of legality and purposefulness (not applicable to regulatory
bodies);
Specifying the internal structure and governance regime;
Approving budget.
Malta
Towards all subordinated bodies:
Giving directions and setting targets, except for the issues where department is obliged by law to act independently;
Monitoring and assessing performance in relation to these directions and targets;
Ensuring timely, effective, efficient and economic performance of functions and delivery of public services.
Poland
Towards all subordinated bodies (unless the special law determines otherwise):
Submits the proposal of the statute of the subordinated body for the approval of the Prime Minister;
Establishing and abolishing subordinated bodies;
Appointing and dismissing heads of the subordinated bodies;
GOV/SIGMA(2021)3 37
Unclassified
Inspecting and supervising their activities;
Issuing binding policy guidelines that may not relate to the issues resolved in the course of administrative proceedings;
In some cases, acting as appeal body in administrative proceedings.
Slovenia
Towards bodies within ministries:
Appointment and dismissal of head (government at the proposal of the minister);
Adopting budget and annual plan of the relevant body upon proposal of its head;
Determining the internal organisation of the relevant body at the proposal of the head (which in practice means approving
the proposal of the head);
Issuing general guidelines;
Issuing binding instructions and orders to undertake specific actions;
Representing the relevant body before the Government and Parliament;
Overseeing the work of the body;
Reviewing the annual report;
Requesting information, documents and reports;
Acting as appeal body in administrative proceedings.
Towards public agencies:
Appointing members of the board (government at proposal of the minister); for some regulatory bodies (electronic
communication, energy) the board is appointed by the Parliament at the proposal of government;
Monitoring legality, efficiency and effectiveness of the work of the relevant agency (excluded for regulatory bodies);
Agreeing on the annual budget of the agency;
Acting as appeal body in administrative proceedings (excluded for regulatory bodies);
Approving loan agreements;
Approving allocation of surplus of revenues over expenses.
Source: The 2005 Government of the Republic Act (Estonia); The 2019 Public Administration Act (Malta); The 1996 Law on the Council of
Ministers (Poland); The 2002 Law on Public Administration and the 2002 Law on Public Agencies (Slovenia)
Note: Instruments of ministerial steering established in the framework laws on public administration might be modified or deactivated with regard to individual bodies by special legislation regulating their status.
A slightly different approach is taken in Ireland, where the ministerial governance frameworks are not
established in detail by law, but are subject to oversight agreements between the ministry and the agency,
allowing for some flexibility and more tailored arrangements. However, even in this case, key elements of
the governance framework are specified at the central level in the Code of Practice for Governance of
State Bodies. This document establishes a catalogue of management mechanisms and lists the issues to
be regulated in detail through oversight agreements between ministries and agencies under their aegis.
Figure 5. Building blocks of the ministerial governance framework for agencies. The case of oversight agreements in Ireland
Clearly defined roles and responsibilities in line with governing
legislation underpinning agency
Level of compliance with Code of Practice for the Governance of State
Bodies
Alignment of Statement of Strategy with portfolio ministry’s Statement of
Strategy
Performance Delivery Agreements
(annual and multi-annual objectives and targets specifying expected level of
performance of agency)
Periodic Critical Reviews
(analysis of the need for continuation of the agency as a standalone body and possible
measures improving its accountability, efficiency and effectiveness)
Remuneration and Superannuation
(rules regarding salaries and other forms of compensation established in line with the
government standards)
Source: Department of Public Expenditure and Reform 2016
The catalogue of specific ministerial managementmechanisms is similar across the European countries. The most important criterion differentiating the scope of ministerial interference in an agency’s autonomy is the type of agency. Some instruments might be deactivated for more
38 GOV/SIGMA(2021)3
Unclassified
autonomous agencies. There is no uniform standard for determining the allocation of individual bodies to the relevant type of agencies. However, in the context of the European Administrative Space (EU member countries and those harmonising their legislation with EU law), the crucial factors to consider are the autonomy standards formulated in the EU acquis for specific type of agencies, especially NRAs. The governance framework for agencies should recognise their special status in some dimensions, while ensuring that they are part of the government administration and remain subject to an effective, results-oriented accountability regime.
Performance management systems
Irrespective of the type of agency, results-oriented performance management, combining agencies’
operational autonomy with accountability for the outcomes delivered, is the cornerstone of the ministry-
agency relationship. Agencies should be accountable to ministries for delivering clearly-defined (agreed)
objectives and targets. Various models exist for such performance management. In international practice,
we may distinguish two models of results-oriented performance management of agencies:
Uniform and strictly regulated model. This model relies on detailed regulation of the whole
process of planning, performance measurement and assessment. Objectives, indicators and
targets are presented in a structured manner, using a standardised performance matrix.
Communication between ministry and agency is regulated by law with strict division of tasks in the
process, leaving little room for negotiation or deliberation on the plans and performance of
agencies.
Flexible and decentralised model. This model is characterised by a modest legislative
framework and a greater role for well-established administrative practice, as well as a more
flexible, sometimes contractual model of relations between the ministry and the agency. It also
concentrates less on the rigorous application of the concept of management by objectives. The
objectives are not always accompanied with measurable indicators and targets, but focused on
setting priorities and assignments of particular relevance.
Both models have their advantages and disadvantages. A flexible and decentralised model may be difficult
to implement in administrations without a long-standing tradition of “performance culture”. The latter model
may look bureaucratic and rigid, but it contributes to greater consistency across the government and helps
to implement basic rules of results-based management to institutions unfamiliar with this model. The former
model was implemented by Portugal, where the performance management system and documents
produced within this process are regulated in detail by law (box 6).
Box 7. Integrated System of Management and Performance Assessment in Public
Administration (SIADAP) in Portugal
The SIADAP was introduced in 2004 as a comprehensive management and performance appraisal system consisting of three components: (1) System for performance assessment of administrative bodies (SIADAP 1); (2) System for performance appraisal of managers in public administration (SIADAP 2); and (3) System for performance appraisal of employees in public administration bodies (SIADAP 3). Each system is regulated in detail by law and Government sublegal acts. In the context of ministry-agency relations, SIADAP 1 is the most important element of the system. It embraces a full cycle of results-based performance management, from setting mission-based, multi-annual strategic objectives for each body, through establishing annual objectives, performance indicators and targets, to assessing their accomplishment and identifying any divergences in meeting them, as well as their causes. Objectives are proposed by each institution, but they are subject to approval by the portfolio ministry. Performance assessment begins with a self-assessment completed
GOV/SIGMA(2021)3 39
Unclassified
by the respective institution. The results are presented in the annual report of the institution. In addition to analysis of the implementation of the objectives, the annual report should include user appraisals of the services provided by the institution, explanation of any failures in delivering expected results, measures planned for improving performance and comparisons with performance of similar services, both at national and international level. The self-assessment is followed by a critical analysis provided by the unit of the portfolio ministry responsible for planning, strategy and assessment. The ministry is not only required to review the performance of individual subordinated bodies, but also to compare performance across all bodies under its supervision and distinguish the best and the weakest performers. Based on the self-assessment and review conducted by the portfolio ministry, the process is concluded with the final score proposed by the head of the institution and approved by the portfolio minister: a) Good performance – all objectives were attained and performance with regard to some objectives was higher than expected; b) Satisfactory performance – all or at least most relevant objectives have been achieved; and c) Insufficient/poor performance – most relevant objectives were not achieved.
Source: Law no. 66-B/2007 of 28 December 2007 establishing the integrated system for management and performance assessment in Public Administration; (Madureira, Rando and Ferraz, 2020[54])
Several other countries follow a more flexible approach. For example, governing agencies through annual
instruction letters (letters of appropriation) is characteristic to Norway and Sweden, while performance
contracts are the key tool regulating ministry-agency relations in Finland or Ireland. These countries,
having extensive traditions of agencification and agency autonomy, opted for less rigorous approaches to
performance management. In particular, while following the general principle of results-oriented
management, they do not require agreement on all of the agencies’ objectives and detailed quantitative
targets for each of them. They also provide greater room for adjusting the general rules of the performance
management process to individual preferences and the practices of the respective ministries and agencies.
Box 8. Governing agencies through annual letters of appropriations and performance contracts
(Sweden, Norway, Ireland, Finland)
Sweden is well known for its long tradition of public agencies (dating back to at least the nineteenth century) and its consistently-applied model of small ministries responsible solely for policy making accompanied by a large number of agencies performing all operational functions. The extensive autonomy of agencies guaranteed at the constitutional level is also characteristic to this model. Since the 1970s, the governance model for the agencies gradually evolved from detailed regulations towards management by objectives. The central managementinstruments are annual letters of appropriations (regeringsbrev) addressed by the Government to each agency. Draft letters are prepared by the relevant ministries, but adopted by the whole Government. The format of the letters is uniform. They usually consist of the following elements:
A list of major objectives and requirements for how their implementation will be reported to the Government – it is interesting to note that the formulation of objectives rarely meets the well-known SMART criteria (specific, measurable, attainable, relevant, time-bound), they are often vague and general.
Other reporting requirements, e.g. additional reports on implementation of specific strategies or legislative acts.
Other specific assignments (e.g. request to present expenditure forecasts for upcoming years) and areas to be prioritised in the agency’s work.
Overall funding provided to agency.
Conditions tied to allocated funds (e.g. indication of funds earmarked for specific projects);
Other financial conditions (e.g. loan limit).
Estimated budget for fee-based operations in which the revenue is available for allocation.
40 GOV/SIGMA(2021)3
Unclassified
The annual reports of the agencies, and all other reports submitted to the Government according to the requirements set in the letters of appropriation, serve as a basis for assessment of the agencies’ performance that takes the form of a performance dialogue between agency and portfolio ministry.
A similar model exists in Norway, but the letters of appropriation (tildelingsbrev) to the agencies are issued by the portfolio ministries. The structure of this document is similar to the one in neighbouring Sweden, yet some differences are worth noting. For example, Norwegian letters of appropriation contain a management calendar envisaging key events in the ministry-agency relationship (e.g. meetings to discuss performance or key steps in the budgetary process for the next year). They also appear to put more emphasis on setting measurable performance indicators for each of the main objectives.
Public agencies in Ireland are governed by their respective portfolio ministries primarily through performance delivery agreements (PDAs), containing both annual and multi-annual objectives and targets, based on the ministry’s statement of strategy. The format and content of the PDA is determined by the Code of Practice for the Governance of State Bodies published by the Department of Public Expenditure and Reform (responsible for public administration matters). The PDA should consist of the key priorities and objectives of the agency aligned with the Government’s strategic policy framework, services and outputs to be delivered by the agency and the resources allocated to attain them, potential risk factors, mechanisms of performance measurement, monitoring arrangements, rules for amending targets (in exceptional cases) and the duration of the agreement (in principle, three years).
Relations between ministries and subordinated agencies in Finland are regulated by the multi-annual performance agreements. These documents contain performance targets for each budget year combined with the allocation of the resources needed to attain them. Performance agreements are developed in negotiations between the ministries and agencies, based on a draft that could be prepared by either party. Drawing from the Government programme, the initial version of the performance agreements includes performance targets for the first year and provisional targets for the second, third and four year. Subsequently, the agreement is updated with specific targets for each of the remaining years. Every year, the agencies report on the implementation of the targets. Further, there is a practice of in-year monitoring of progress in delivering the agency’s objectives, through mid-term reports every six months and regular dialogue between ministries and agencies. At the end of the four-year period covered by the agreement, a detailed performance report is prepared.
Source: (Jann et al., 2008[55]) (Levin, 2009[56]) (Askim, 2019[57]) (Öberg and Wockelberg, 2020[58]);review of annual letters of appropriation of various agencies, (Department of Public Expenditure and Reform, 2016[27]) (Ministry of Finance of Finland, 2020[59])
Regardless of the model selected, a performance management regime should embrace all public
agencies, though with some minor modifications for NRAs. In essence, the portfolio ministries should
abstain from imposing specific objectives and targets on agencies, but rather seek agreement on major
priorities. The performance of NRAs should be discussed with the portfolio ministry, but any rating or
imposing of sanctions for unsatisfactory performance should be avoided.
Classic ministerial controls
The performance management system is the foundation of the ministry-agency relationship, determining
the scope of the agency’s autonomy and accountability. However, ministries also possess a range of
classic control instruments. The catalogue of tools of ministerial control may differ, especially considering
the type of agency. The most crucial restrictions apply to the NRAs. The matrix below provides a list of
typical tools with a special focus on tools that should be excluded for NRAs, according to the standards
established by the EU acquis.
GOV/SIGMA(2021)3 41
Unclassified
Figure 6. Classic mechanisms of control of agencies by portfolio ministries
GOVERNANCE CORE FUNCTIONS ORGANISATIONAL MATTERS
Appointment and dismissal of the senior management
Developing policies in the area of agency’s
activities
Monitoring agency’s compliance with the regulations on internal management through audits
and inspections
Approving the internal organisational
structure
Setting general policy guidelines Approving or agreeing on the budget in
co-operation with the ministry of finance
Creating or terminating territorial branches Requesting information, documents and
explanations Approving or determining the staff number
Issuing detailed instructions and orders on
specific activities
Reviewing administrative acts issued by the
agency
Notes: Dark blue represents tools applicable to all public agencies. Light blue represent tools that may not be applicable to NRAs.
The administrative capacity of parent ministries is often neglected in discussions on the agency setup.
However, even when a law or a contract introduces a sufficient management mechanisms, only institutions
with sufficient capacity can effectively use them to prevent the risk of losing political control, without
violating the agency autonomy. As early as 2001, SIGMA warned that trained staff, adequate information
systems or sufficient financial resources are needed to carry out control and governance arrangements
embodied in legislation; an inability to implement them invites failure (OECD, 2001[12]).
Staffing and information can become an issue even in countries with long traditions of constructing ministry-
agency relations. New skills and competencies of parent ministry staff are needed, in particular when
hierarchical relations (based on traditional HR and financial controls) are supplemented with or replaced
by a results-oriented approach. Some countries have developed specific competency frameworks for civil
servants working in this role (Cabinet Office, 2014[60]), introduced educational and training programmes
(Cabinet Office, 2014[61]) or reallocated staff in order to address supervision capacity problems (OECD,
2015[62]). Some administrations have issued central guidance on the supervision of executive agencies by
ministries (Department of Public Expenditure and Reform, 2016[27]) or guidance for members of the boards
of such bodies (OECD, 2018[63]).
Proper internal arrangements need to be made where financial supervision and that of policy execution
are formally attributed to different units in the ministry, to prevent inconsistent expectations for the
subordinated agency (OECD, 2015[62]). In cases where a representative of a ministry is a member of an
agency board, the oversight role in the ministry should not be the responsibility of the same individual.
Developing a culture of ministry-agency collaboration and clarifying roles
Day-to-day relations between (portfolio) ministries and agencies depend not only on formal mechanisms,
but also on something much less tangible, namely a culture of co-operation based on shared goals and
values and a common understanding of the roles, responsibilities and autonomy of both actors. Particularly
in the Western Balkans and the European Neighbourhood, where the concept of autonomous agencies
remains a novelty, it is important to invest time and resources to raise awareness among portfolio ministries
and agencies about the specific nature of the ministry-agency relationship. Building a culture of
independence within an autonomous agency and in its ecosystem and relations with other government
actors takes time. (OECD, 2017[47])
42 GOV/SIGMA(2021)3
Unclassified
Building a collaborative culture is more difficult than establishing formal rules, but may be supported with
general guidelines. For instance, the UK Government developed a Code of Good Practice for partnerships
between departments (ministries) and arm’s length bodies (agencies).
Box 8. The UK Code of Good Practice for partnerships between departments (ministries) and
arm’s length bodies (agencies)
The partnership between ministries and agencies should rely on four principles: (1) Purpose – mutual
understanding of the purpose, objectives and roles of agencies set out in the relevant documents, clear
lines of accountability; (2) Assurance – proportionate approach of the ministry to supervision of agencies,
giving them autonomy to deliver effectively and ensuring that their performance is assessed by the
ministry; (3) Value – sharing skills and experience between agency and ministry; and (4) Engagement –
open, honest, constructive and trust-based relationship based on clarity about mutual expectations.
In addition to these general principles, more specific guidelines are also formulated, including:
There is a strategic alignment between the purpose and objectives of the ministry and the agency.
The ministry’s approach to supervision (assurance) of the agency is based on an assessment of the risks posed by the agency.
The ministry has an appropriate overview of operations of the agency, proportionate to its purpose and required degree of autonomy.
The ministry and agency have access to the data they need to assess the agency’s performance;
There is a regular exchange of skills and experience between ministry and agency, including secondments, joint programmes or project boards, forums for staff of both bodies to learn from each other.
There is a clear process to resolve disputes between ministry and agency.
The relationship between ministry and agency is regularly reviewed and assessed by both parties.
Source: (Cabinet Office, 2017[64])
An important element of the culture of ministry-agency relationships is clarity about the roles of each body.
The policy development role is crucial. In principle, the ministry should be responsible for policy formulation,
while the agency for policy implementation is under ministerial oversight. However, practice in countries
with an extensive track record in agency governance demonstrates that this clear division of labour may
become blurred and agencies may also have substantial impact on the formulation of the public policies
they are expected to implement. For example, as agencies in Sweden have more expertise and capacities
(especially staff), they act as partners to the relevant ministries in policy design (Niklasson and Pierre,
2012[65]). In Germany, in the context of the global financial crisis of 2008, the financial market regulatory
agency took a leading role in shaping the policy response with regard to financial market policies (Handke,
2012[66]). Upholding the separation between policy making and policy implementation functions through
agencification does not always work in practice. However, instead of trying to prevent agencies from
influencing public policies, the government should rather clarify the rules of such involvement. Based on
international experience, the following instruments could be considered:
Ensuring that the role of agencies in developing policy or legislative proposals is subsidiary and
the respective unit and officials of the ministry remain in command.
GOV/SIGMA(2021)3 43
Unclassified
The influence of the agency is focused on providing data, insights from practice of implementation
of policies and laws and comments on the ministerial proposals from the perspective of an
organisation that has more insight into day-to-day practice in the respective field.
Policy making is not “outsourced” by the portfolio ministry to the agency by requests for developing
full policy or legislative proposals.
Any proposals developed by the agency are reviewed by the respective unit and officials in the
ministry before they are taken over as official government policy proposals.
The co-operation between ministry and agency in the policy making process could be
institutionalised in the form of working groups or taskforces.
The expectations of the ministry in terms of agency involvement in policy making should be clearly
established in the performance management framework.
Nevertheless, direct agency involvement in policy making should be reserved for exceptional cases.
Portfolio ministries have the democratic legitimacy and responsibility to develop public policies and the
primary role of agencies is ensuring that these policies are effectively implemented.
Summary - Good practices and tools for agency governance
There is no ‘one-size-fits-all’ solution for managing agencies. However, there is a common list of challenges
requiring similar actions. Based on a review of international trends, practices and standards, a set of good
practices and tools can be identified that help governments manage the agency landscape and strike a
balance between agency autonomy and accountability. These good practices and tools for agency
governance are not widely applied in the Western Balkans and the European Neighbourhood, as shown
in the next Chapter. In transition economies, the elements above often need to be specified in the legal
and institutional framework for public agencies and overall responsibility for managing the government’s
policy towards agencies should be clearly allocated. Once the legal and institutional foundations are set,
the quality and consistency of day-to-day management, as well as the stability and continuity of high-level
commitment to good governance of agencies, will be the key factors determining the performance of the
public administration.
44 GOV/SIGMA(2021)3
Unclassified
Figure 7. Summary of good practices and tools for agency governance
WHOLE-OF-
GOVERNMENT
LEVEL FUNCTIONS
Consistent standards for
governing agencies and
ensuring regular
optimisation of the
agency landscape
Taxonomy (typology) of agencies (established in the legislative framework) is
based on different levels of autonomy required by them to perform their
functions, taking into account the EU standards for specific types of agencies.
Procedure for ex ante review of proposals for creation of new agencies.
Regular reviews of existing agencies at individual level to establish the case for
their continuation and at collective level to seek optimisation e.g. by grouping
agencies or considering alternative delivery options for government functions.
Whole-of-government regulations and standards applicable to all agencies
concentrated on ensuring consistent and efficient management across
government.
PORTFOLIO
MINISTRY LEVEL
FUNCTIONS
Mechanisms for agency
accountability for results
preserving necessary
autonomy
Steering framework structured around results-oriented performance
management system and ensuring that agency objectives and targets are
agreed (negotiated) between portfolio ministry and agency.
Clear delimitation of the agency’s autonomy, including “no-go zone” for the
portfolio ministry.
Classic mechanisms of ministerial control, such as appointment of the agency’s
management, conducting inspections, requesting information or issuing
guidelines and instructions (see box 6 for NRA-specific mechanisms).
GOV/SIGMA(2021)3 45
Unclassified
Introduction
This Chapter begins by briefly explaining how the central government landscape has changed in the
Western Balkans and the European Neighbourhood since the idea of agencification became established,
identifying the main causes of what may be characterised as “agencification on steroids” and five
challenges that this development has resulted in for the public administration and citizens. The following
section analyses the levels of autonomy and accountability of different types of agencies. The final section
summarises lessons learned from past reform initiatives.
Ten administrations were analysed, drawing primarily on data from SIGMA Monitoring Reports and a
survey. The survey covered: Albania (ALB), Armenia (ARM), Bosnia and Herzegovina (BIH), Georgia
(GEO), Kosovo (XKX), Moldova (MDA), Montenegro (MNE), North Macedonia (MKD), Serbia (SRB) and
Ukraine (UKR). Details about the survey are presented in the next section. The Western Balkan
administrations benefited from two full rounds of SIGMA monitoring (2017 and 2021), whereas in Ukraine
and Armenia only a baseline was established (2018). No SIGMA Monitoring Reports are available for
Georgia and Moldova in this area, but data has been collected for this paper for a subset of sub-indicators
to enable comparison6. SIGMA’s engagement in agency governance reforms in Albania, Kosovo, North
Macedonia, Serbia and Ukraine also provided useful additional information to validate the other data
sources and to write the final sub-section of this chapter.
Agencification in the Western Balkans and the European Neighbourhood – an
overview
The global phenomenon of agencification was enthusiastically embraced by all Western Balkan and
European Neighbourhood administrations, following the general trend in other post socialist countries of
Central and Eastern Europe (CEE). Governments launched very ambitious reform programmes to
restructure central government and public agencies – often driven by EU accession – but had much less
“reform capacity” than the EU and OECD countries that were the source of inspiration. Previous scholars
have documented the warped logic that the wave of agencification brought in CEE countries’ transition
2. Uncontrolled proliferation of agencies and lack of mechanisms preventing unjustified
creation of new bodies.
There are no examples of comprehensive and robust policy and regulatory frameworks that can prevent
uncontrolled proliferation of agencies in the administrations covered by this study. No administration fulfills
all of the criteria from the Methodological Framework for the Principles of Public Administration. However,
some have made progress since 2017 (table below). All administrations now have in place procedures for
establishing, merging and abolishing central government bodies, but North Macedonia is the only one with
an overall policy plan for institutional development of the central government. Albania is the only
administration with a body within central government that has formal responsibility for regular reviews of
the organisation of central government (though having a weak record on performing this function in
practice), and fulfills all criteria except having an overall plan for institutional development. Despite the
existence of policies and regulations to manage the process of creation of new bodies, the proliferation of
agencies continues without strong ex ante control (see figure 9 below).
Table 8 Existence of policies and regulations to manage central government agencies
Criteria ALB ARM BIH XKX MNE MKD SRB UKR
Plan for institutional development of central government is specified in policy document(s)
▲ ▲
Procedure for establishing, merging and abolishing each type of central government body is specified in the legislation
● ● ●
▲
▲
▲ ● ●
Procedure for establishing, merging and abolishing each type of central government body requires participation of prime minister’s office, ministry of finance and HRM authority
▲
▲ ▲
▲
▲
Creation of a new body must be accompanied by ex ante analysis covering at least: 1) assessment of the need to create the new body; 2) analysis of alternatives to creation of the new body; and 3) estimated cost and staffing of the new body
▲
▲
▲ ●
▲
▲
A body within central government is responsible for regular reviews of organisation of central government and planning institutional development
▲ ▲
▲
Source: (OECD, 2018[68])
GOV/SIGMA(2021)3 49
Unclassified
Note: Data for the six Western Balkan administrations is from 2021. Data for Ukraine and Armenia is only available for 2018. A green triangle
denotes that this criterion was not fulfilled in 2017 but was in 2021. A red triangle shows where the criterion was fulfilled in 2017 but no longer
in 2021.
3. Lack of clear and comprehensive typology of agencies setting consistent and
common rules for their autonomy and supervision.
The framework laws for organisation of public administration recognise agencies as a part of the public
administration in all cases. The official typologies of agencies differ considerably but they follow a similar
model. While most of them clarify the legal status of public administration bodies and set the basic rules
for their internal organisation (managing bodies, institutional locus), they lack clear distinction between
various types of bodies, leading to different degrees of autonomy. In some cases, such as in Kosovo, even
if a framework law has been adopted by parliament, the positive effects of an official typology are nullified
by the existence of conflicting sector legislation and lack of harmonisation of laws. In Albania, the distinction
in the law is so unclear that most of public administration bodies meet the criteria for each type of body.
Moreover, bodies of the same type may operate in different regimes relating to the financial management
or status of the staff. Therefore, in practice, most agencies continue to operate under individually crafted
governance schemes resulting more from political bargains rather than a clear, overarching vision of
organisation of the public administration. Across the administrations covered, many public bodies continue
to exist outside the official typology, enjoying “special” status. All of these factors result in a chaotic,
fragmented and unmanageable organisational landscape, with very practical and tangible deficiencies,
such as unexplainable differences in salaries for the same position and other variations in employment
conditions across the public sector.
Table 9 Clarity and comprehensiveness of official typology of central government agencies
Criteria ALB ARM BIH XKX MNE MKD SRB UKR
Legal status is explicitly regulated for all types of central government bodies
● ●
▲
▲ ● ●
Functional criteria for establishment are explicitly regulated for all types of central government bodies
▲
Managing bodies are explicitly regulated for all types of central government bodies
● ● ● ● ● ●
Subordination/supervision schemes are explicitly regulated for all types of central government bodies
● ● ▲ ▲
●
Degree of autonomy in financial management and HRM is explicitly regulated for all types of central government bodies
● ●
▲ ● ●
Source: (OECD, 2018[68])
Note: Data for the six Western Balkan administrations is from 2021. Data for Ukraine and Armenia is only available for 2018. A green triangle
denotes that this criterion was not fulfilled in 2017 but was in 2021. A red triangle shows where the criterion was fulfilled in 2017 but no longer
in 2021.
4. Large number of classical executive agencies placed under (weak) stewardship of
parliaments and thereby escaping oversight and ministerial management.
The notion of “excessive autonomy”, especially characteristic in the Western Balkans, is the trend to set
up agencies under parliament that would normally be part of the government administration and carry out
traditional executive functions, such as implementation of laws and policies or even service delivery. This
often leads to duplication of agencies, tasks, waste of resources, problems with policy co-ordination and
undermines the government’s legitimate right to oversee agencies and policy implementation and to control
any irregularities in the use of public funds.
50 GOV/SIGMA(2021)3
Unclassified
As described in Chapter 3, bodies performing executive functions of the state should – as a rule - remain
part of the government administration, reporting to respective portfolio ministries. This allows governments
to oversee implementation of policies and ensure consistent activities of the whole administrative
apparatus. The role of the legislature is to exert classic parliamentary oversight, e.g. request information
and conduct inquiries or commission audits of relevant government bodies. Parliaments do not have
capacities to manage agencies and conduct effective day-to-day oversight of their activities. Paradoxically,
parliaments’ oversight capacities appear weakest where the number of bodies reporting to the legislature
is the highest, as shown in Figure 7. In Kosovo, which has the highest number of agencies reporting to the
parliament except for Bosnia and Herzegovina, only one person in the Assembly is tasked with overseeing
reporting and daily management of agencies.9 Neither the members of legislature nor the small secretariats
have time or sufficient sectoral expertise to effectively perform their supervisory functions. It is simply not
their job. Thus, agencies reporting only to the parliament often operate in a governance vacuum, as largely
self-governing bodies. They are not integrated into the government’s policy processes and the
government’s steering mechanisms. With very limited parliamentary oversight and unclear position in the
state administration, they are also particularly exposed to the risk of agency capture by the interest groups.
Figure 9. Number of agencies subordinated to the parliament in the Western Balkans and the European Neighbourhood (constitutional bodies excluded)
Source: (OECD, 2018[68])
9 SIGMA Monitoring Report 2017 for Kosovo, page 87
23
22
21
19
18
10
10
3
2
0
Bosnia and Herzegovina
Kosovo
Serbia
North Macedonia
Albania
Montenegro
Moldova
Georgia
Armenia
Ukraine
GOV/SIGMA(2021)3 51
Unclassified
Note: Data for the six Western Balkan administrations is from 2021. Data for Ukraine and Armenia is from 2018. Moldova data was collected by
SIGMA in 2021 for the purposes of this paper. For Bosnia and Herzegovina, 3 agencies are in the Federation of BiH, 10 are in the Republika
Srpska, 2 in the Brčko District and 8 at State level. Dark blue represents the Western Balkan region and light blue the Neighbourhood region.
Bodies mentioned in the constitution, such as the supreme audit institution, the ombudsperson institution, central bank, etc. are not included.
The exclusion of constitutional bodies distorts the number for Ukraine, as the Ukrainian Constitution stipulates that some bodies performing
executive functions are subordinated to the Parliament. This relates, for example, to the competition protection authority and the State Property
Fund.
As mentioned in Chapter 2, subordination of agencies to the parliament is not required by the EU acquis,
even for the most autonomous NRAs. The dominant model among the EU Member States ensures the
functional autonomy of NRAs within their setting in the government administration. In the Western Balkans
and the European Neighbourhood the situation is different, as formal subordination of the NRAs under the
parliament of often simplistically and incorrectly equated with “independence”.
The table below presents data for four types of NRAs, responsible for competition protection, energy
markets, audio-visual media services and electronic communications. It shows that in the Western Balkans
and the European Neighbourhood the vast majority of the NRAs report directly to national parliaments, not
the government administration. This clearly contrasts with the dominant practice of EU Member States,
where only a few NRAs are subordinated to the legislature (OECD, 2021[69]). Audio-visual media service
regulators are an exception, as close to a majority of EU NRAs in this sector report to parliaments.
Nevertheless, the general trend is clear. The Western Balkans and European Neighbourhood provide
examples of regulatory “gold-plating”, i.e. excessive transposition of the EU rules, going far beyond what
is required by the EU acquis.
Table 10 National regulatory authorities reporting to the Parliament in the Western Balkans and the European Neighbourhood
Regulators ALB ARM BIH GEO XKX MDA MNE MKD SRB UKR Total (10) EU27
Competition protection ● ● ● ● ● ● ● 7 5
Energy markets ● ● ● ● ● ● ● ● ● 9 6
Audio-visual media ● ● ● ● ● ● ● ● 8 12
Electronic communications
● ● ● ● ● 5 5
Source: Data collected by SIGMA based on desk review of legal acts.
Note: A NRA is considered as reporting to the parliament when its management is appointed and dismissed by the legislature.
5. Lack of results-based management and performance evaluation
At the level of ministry-agency relations, the principal problem is a lack of results-based management of
agencies and a culture of performance evaluation with strong mechanisms. Annual and/or multi-annual
plans often do not have specific objectives and targets to be achieved by the agencies. Performance
monitoring is rarely carried out and there are few established practices for providing regular feedback about
agency performance (figure below).
52 GOV/SIGMA(2021)3
Unclassified
Figure 10. Results-oriented governance of agencies in the Western Balkans and the European Neighbourhood
The annual plan of the subordinated body
contains specific objectives and measurable
targets approved by the ministry or agreed
between the ministry and the subordinated
body
Progress towards objectives is monitored
by a relevant unit of the ministry, at least
annually
The last annual report contained
information on the level of outcomes
against predefined objectives and targets,
and the ministry provided feedback on this
in writing
Source: (OECD, 2018[68])
Note: Data for Georgia and Moldova not available.
How autonomous, how accountable?
As shown in Chapter 3, good governance of agencies entails a balance between the necessary level of
autonomy for the agency and the necessary accountability mechanisms towards the portfolio ministry for
the outcomes achieved. Autonomy is not a binary concept, although this is often how it is perceived in the
Western Balkans and the European Neighbourhood. Different types of agencies require different levels of
autonomy across different dimensions, depending on their functions.
To explore the levels of autonomy of different types of agencies in the Western Balkans and the European
Neighbourhood, SIGMA conducted a survey among agencies and portfolio ministries. This provides unique
insights into the nature and specific challenges of agencification in the Western Balkans and the European
Neighbourhood. This data enables us to understand the five challenges mentioned previously in more
detail and provide better evidence-based advice on the necessary reforms.
With a total of 236 respondents, this survey provides the first and most comprehensive cross-country
empirical evidence for the state of agency governance in the region. More than 136 senior managers of
public agencies and 100 senior civil servants in portfolio ministries were surveyed. A total of 17 sectors
were covered. Agencies operating in the following 4 sectors are collectively referred to as NRAs:
audio-visual media services
competition protection
Yes22%
No78%
Yes10%
No90%
Yes24%
No76%
GOV/SIGMA(2021)3 53
Unclassified
energy markets
electronic communications.
The remaining sectors covered are civil aviation, civil aviation safety investigations, civil registration, data
protection, financial markets, health insurance, pension funds, national statistics, prison administration,
road administration, tax administration, food safety inspections and labour inspections.
How autonomous?
Both respondents from agencies and portfolio ministries recognise that public sector agencies in general
have extensive autonomy in the regions. Respondents from NRAs (especially competition protection
authorities and audio-visual media services regulators) as expected report the highest degree of autonomy.
Respondents from other public agencies report lower levels of autonomy, but 40% state that they have
“extensive autonomy". Overall, this corresponds with the SIGMA findings above that agencies have
extensive autonomy in the regions and the fact that a vast majority of the NRAs in the Western Balkans
and the European Neighbourhood report only to parliaments.
Table 11 Agency autonomy – differences in views across agencies and portfolio ministries
Average (1-5) 1
Little autonomy
2 3 4 5
Extensive autonomy
Senior managers in all agencies 3.85 3% 9% 27% 20% 40%
4) Implementation is key, and hinges on the ability of the administration to manage a whole-of-
government reform agenda. Overambitious restructuring programmes can overwhelm the
administration and may be abandoned. Lengthy analytical exercises risk missing the window of
opportunity when political support is mobilised. Sequencing reforms in concrete, management
steps in action plans is recommended.
5) There may be strong resistance to changes from various interest groups. The fragmented
institutional landscape with a multitude of agencies not accountable to ministries has powerful
advocates in each administration, e.g. among the management of agencies enjoying excessive
benefits or autonomy, or among political patrons of some agencies. Even donor organisations
lobby to keep the agencies that they helped create. Misquoting the EU acquis is a common tactic
to halt reforms.
6) The advantages of a reform should be clearly explained to the political leadership, in particular the
political gain obtained by finding a proper balance between autonomy and accountability and
creating a setup where performance objectives for the agencies are co-created by the government
and the agencies can be held accountable for performance.
Summary
Combining the findings of SIGMA monitoring, the outcomes of reform initiatives and the insights from the
survey carried out for this paper, provides for a consistent analysis of where the Western Balkan and
European Neighbourhood administrations are in their journey towards good governance of agencies. All
in all, the notion of a “governance vacuum” may serve as the most accurate depiction of agency
governance in the Western Balkans and the European Neighbourhood. This vacuum begins at the macro
level with the lack of a clear vision and logic for when agencies should be created, which type of agency
should be applied in specific cases, how to prevent uncontrolled proliferation and how to ensure that they
are governed according to clear and consistent rules. It continues at the meso-level with a lack of
accountability mechanisms in the interaction between portfolio ministries and public agencies.
As demonstrated by the survey, the governance vacuum at the central level is not compensated by
effective governance of agencies at the level of individual portfolio ministries, and it is not a matter of
legalistic versus results-oriented management culture. The dominant model of relations between portfolio
ministries and agencies could be described as a mixture of rather extensive autonomy and a more general
absence of performance monitoring and accountability for results. Portfolio ministries do not apply the
typical tools for managing agencies developed by OECD countries. The Western Balkans and the
European Neighbourhood have a unique and arguably dysfunctional characteristic in largely bypassing
senior civil servants and placing the formal management and oversight responsibility directly in the hands
of ministers.
The consequence is that portfolio ministries fail to have full ownership of their policy areas and lack
sufficient interest in enhancing the agencies’ performance. In the case of the NRAs, there is also an issue
of misunderstandings about the requirements of the EU law. The table below summarises the key problems
and deficits identified both at the level of the whole system and individual ministry-agency relations, as well
as the consequences of these problems for the overall performance of government administration.
68 GOV/SIGMA(2021)3
Unclassified
Table 13 Summary of key deficits in agency governance in the Western Balkans and the European Neighbourhood
LEVEL MISSING ELEMENTS EFFECTS
Macro-level
(government policy and general
legislative framework)
Sound typology of agencies and clear criteria for establishment and selection of the adequate
type
Rigorous ex ante control of agencification and
analysis of alternative options for delivering
government functions
Proliferation of agencies lacking clear justification and
benefits;
Steering problems associated with multitude of agencies of various types, degrees of autonomy and
governance regime
Regular reviews and fine tuning of the agencies’
landscape
Inefficient administrative structures and missed
opportunities for efficiency gains and savings
Consistent governance regime covering all agencies and striking proper balance between agencies’ autonomy and accountability,
including preventing establishing agencies
directly under supervision of parliaments
Autonomy of agencies determined individually for each agency resulting often in their excessive autonomy or special privileges, e.g. exemption from rules on
transparent recruitment, salary scales or financial
controls;
Strong ministerial ownership and accountability for relevant policy areas due to high number of
agencies located outside the executive and
reporting to the parliament only
High risk of inefficiencies and irregularities (e.g. agency capture by interest groups) associated with weak
parliamentary oversight;
Lack of instruments for the government to ensure
consistent implementation of sectoral policies
Meso-level
(ministry-agency relations)
Well-established, results-oriented culture of ministerial steering and performance
management
Agencies operating without clearly defined expectations and rigorous accountability for delivering
policy outcomes
Clear allocation of responsibility for steering
agencies within the ministry
Steering responsibilities are not effectively and consistently performed, leaving the agencies in the
governance vacuum
The agencification experience of the Western Balkans and the European Neighbourhood is already so
vast that it serves as a basis for tailored and specific recommendations. The table below summarises the
principal “dos and don’ts” for governing agencification.
Table 14 Dos and don'ts of agencification - key recommendations
Do’s Don’ts
Consider a wide array of organisational options for delivering government functions and make final choice
based on thorough, unbiased analysis rather than dogmatic,
theoretical assumptions and models;
Remain sceptical towards any proposals for creation of new institutions – demand clear and strong business case for
establishment of new bodies and consider it as a last resort;
Work towards “compact government” – constantly counteract natural tendency of administrative apparatus to
swell by pursuing smaller, more integrated, but also more
efficient and effective administration;
Consider governance of agencies as one of the key responsibilities of ministries and ensure that it is clearly assigned with the necessary management and supervision
capacities;
Search for balance between autonomy and accountability in
governing agencies – too many governments enthusiastically adopted the idea of autonomous agencies, forgetting about securing accountability of agencies. Let the
(agency) managers manage, but also hold them
accountable for the results they deliver.
Do not be dogmatic – the idea of splitting policy making from policy implementation does not serve as a sufficient
justification for massive agencification;
Do not take the benefits of agencification for granted – the
abstract and theoretical assumptions about efficiency gains of agencification are not sufficient to justify the creation of
new bodies;
Do not expect that staff in ministries do not need specialised
skills when tasks are delegated to agencies;
Do not confuse agency autonomy with independence – agencification makes sense if considerable autonomy
(combined with accountability) is provided to newly created agencies. However, even the most autonomous agencies, such as the NRAs, do not constitute a “fourth branch of the
government”. They remain part of the executive that should
be held accountable for implementing government policies;
Do not let the agencies move too far away from the government – agencies should remain firmly located within public administration in terms of staffing, salary or financial
management rules. Too often agencification is used as a vehicle to create a grey zone of excessively paid political
appointees and non-transparent spending.
GOV/SIGMA(2021)3 69
Unclassified
References
Askim, J. (2019), “Quasi-contractual ministerial steering of state agencies: Its intensity, modes,
and how agency characteristics matter”, International Public Management Journal, Vol. 22/3,
pp. 470-498.
[57]
Bach, T., B. Niklasson and M. Painter (2012), “The role of agencies in policy-making”, Policy and
Society, Vol. 31/3, pp. 183-193.
[15]
Baldwin, R., M. Cave and M. Lodge (eds.) (2010), Introduction: Regulation—the Field and the
Developing Agenda, Oxford University Press.
[38]
Balkan Policy Research Group (2020), Public Administration Reform in Kosovo - Constant
Struggle to Make It.
[73]
Beblavý, M. (2002), “Understanding the waves of agencification and the governance problems
they have raised in Central and Eastern European Countries”, OECD Journal on Budgeting,